4. Managing Nodes¶
4.1. Resource Discovery¶
As part of the Robin Platform installation process, resource discovery is run on the node wherein which details about the physical configuration, hardware limits and resource availability are discovered. The purpose of this is two fold. First this process allows Robin to gain a better understanding of the machine in terms of both the compute and storage resources it can provide for application deployment as well as allow Robin to better optimize the nodes usage within the cluster. Moreover it allows Robin to reserve resources that were already in-use pre-Robin installation so as to make sure resources aren’t diverted inappropriately.
The following properties of the node are discovered:
Disks
CPUs
Random Access Memory (RAM)
Network Devices
NUMA Configuration
Details on what is captured with regards to each one of the above aspects alongside how they are captured are described below.
4.1.1. Disk Discovery¶
Robin leverages a multitude of sources to discover the disks that are available to a node. Some of the commands and directories used to attain the below details are: lsblk, partprobe, pvs, blkid and /proc/mounts. The following details are captured for each disk (if present):
Devpath
Capacity
Physical Sector size
WWN (along with make and model)
Media type
This discovery is primarily performed in order determine the storage capacity of the node and consequently the amount of storage resources the host can provide to its assigned resource pool. However not all discovered disks are considered to be viable to be used as storage. Those that are excluded are primarily considered to be root devices. Moreover we do not consider devices that have parititions, file mounts or are LVMs as devices to be used for storage. This is due to the fact that when a storage role is assigned to a host Robin formats devices marked for storage.
4.1.2. CPU Discovery¶
Robin leverages the information stored within the /proc/cpuinfo file in order to discover the CPUs present on a machine. The following details are captured for each CPU (if they are present):
CPU Index
Physical ID
Vendor ID
CPU Clock Speed
Model Name
Cache Size
Core ID
In addition to the above, Robin discovers if a processor is online, and the min/max frequencies it can operate at by utilizing the values stored in the cpufrequency directory. The reason behind this extensive detail gathering is to enable resource tracking via identification for chargeback purposes and resource accounting whilst planning application deployment.
Moreover if there are any isolated CPUs present on the system, Robin will distinguish them from the general CPUs. This is because Robin will only utilize isolated cores, if present, for application deployment due to the increased performance that can be achieved with isolation and CPU pinning. As a result, from Robin’s perspective the host will only have these isolated cores to contribute to the resource pool.
4.1.3. Memory Discovery¶
Robin leverages the information stored within the /proc/meminfo file in order to discover the overall memory statistics of the host as well as certain breakdowns. The following details are captured (if they are present):
Total memory available on the host
Total number of 2M (2048 KB) Hugepages
Total size of 2M (2048 KB) Hugepages
Total number of 1G (1024 MB) Hugepages
Total size of 1G (1024 MB) Hugepages
Breakdown of memory banks including their size, speed, manufacturer and type
Once again the reason behind this extensive detail gathering is to enable resource tracking via identification for chargeback purposes and resource accounting whilst planning application deployment. Moreover these details are necessary to determine how much memory to reserve for use by Robin services and to account for the memory that is already being used by non-Robin services. The latter is especially important as memory is not an overprovisionable resource.
4.1.4. Network Device Discovery¶
Robin leverages the information stored within the /sys/class/net directory in order to discover properties about the physical interfaces present on the node. The following details are captured for each NIC (if they are present):
Name of the NIC
Maximum Transmission Unit
MAC address
Product ID
Vendor ID
Attached CPUs
Domain, bus and slot configuration
The NUMA node to which it is attached if applicable
The number of Virtual Functions present (for SR-IOV NICs)
IP Addresses
IPv4/v6 Gateways
The reason behind this extensive detail gathering is to enable appropriate IP allocation/tracking when deploying pods and ensuring network connectivity between instantiated containers. In addition the details help to validate, plan and orchestrate certain application deployment configurations.
4.1.5. NUMA Configuration Discovery¶
Robin leverages the information stored within the /sys/devices/system/node* directories in order to discover properties and resource limits of each NUMA node present on the host. The following details are captured for each NUMA node (if they are present):
NUMA Node ID
Total memory available
Total size of 1G (1024 MB) Hugepages
Total size of 2M (2048 KB) Hugepages
Associated General CPU IDs
Associated Isolated CPU IDs
Number of allocations present on each associated CPU ID
The reason behind this extensive detail gathering is to ensure that whilst application deployment planning we do not over allocate resources or incorrectly track the used/available resources. This is especially true given the fact that Robin attempts to make every application as NUMA aware as possible in order to boost performance and optimize workloads. As with CPU discovery, if any isolated CPUs are attached to the NUMA node they will be the only processors tracked for the node.
Moreover given all the above is captured alongside details of any SR-IOV Nics attached to particular NUMA node, Robin has the capability to deploy applications completely enclosed within the NUMA boundary if specified as condition during application deployment.
Note
The aforementioned isolated CPUs can be achieved by manually editing the isolcpus within the GRUB configuration for the host which results in the kernel scheduler never utilizing the CPU’s specified.
4.2. Robin Node Roles¶
There are three Robin roles that can be assigned to hosts: Manager, Compute and Storage. Depending on the resources present on a node you can initialize it to have one or more of these roles. This allows for a more granular control over what resources a node can provide for allocation. Below is a description for each role:
The Manager role should be designated to a node which is intended to be part of the Robin Control Plane. The first node which is added as a Manager will be considered the master and have essential Robin services, including the RCM server, file server and event server, running on it. As a result, it will essentially control and manage (hence the name of the role) the agent nodes. This entails handling all external communication with Robin via APIs, maintaining the most current replica of the PostgreSQL database etc. However if a failover occurs, these services are moved and run on the newly elected master. For every Robin cluster there is a maximum of 3 Manager nodes (this is necessary for HA installations).
The Compute role should be designated to a node which is intended to host pods and their respective containers. As a result, these nodes will have resources such as memory and CPU utilized by application deployments. When orchestrating applications the Robin server will try to provision pods on nodes with this roles, and track the aforementioned resource allocations/availability.
The Storage role should be designated to a node which is intended to provide storage, as indicated by its name, for applications deployed on Robin. As a result, any volumes needed for deployed applications will be created and mounted on devices on nodes with this role set.
The aforementioned roles are not mutually exclusive to one another. As a result, a node could have any combination of the three roles (including all of them at one time). However a host within Robin cluster must always have role assigned to it unless it is to be decommissioned.
The following commands are described in this section:
|
Add one or more role(s) to a host |
|
Move one or more role(s) out of maintenance mode for a host |
|
Move one or more role(s) into of maintenance mode for a host |
|
Remove one or more role(s) from a host |
4.2.1. Adding role(s)¶
To add a role to a host that is already registered within the Robin cluster, issue the following command:
# robin host add-role [<hosts>] [<roles>]
--rpool <rpool>
--disks <disks>
|
Comma separated list of hosts to add role to. |
|
Comma separated list of roles. Valid values are ‘storage’, ‘compute’, and ‘manager’ |
|
Assign a resource pool prior to adding a role. |
|
Comma separated list of disk WWNs to add as part of storage role addition. |
Example:
# robin host add-role centos-60-212,centos-60-214 Compute,Storage --wait
Job: 18 Name: HostAddRoles State: VALIDATED Error: 0
Job: 18 Name: HostAddRoles State: WAITING Error: 0
Job: 18 Name: HostAddRoles State: COMPLETED Error: 0
Adds a role to a host that is already registered within the Robin cluster.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: add_roles- This mandatory field within the payload specifies that the add role operation is to be performed.roles: <list_of_roles>- This mandatory field within the payload is a list of roles that should be added to the specified host. Valid values include ‘storage’, ‘compute’, and ‘manager’.drives: <list_of_wwns>- Utilizing this parameter by specifiying a list of WWNs of drives results in the disks associated with the aforementioned WWNs being added alongside the addition of the storage role.rpool: <rpool_name>- Utilizing this parameter by specifying a resource pool name results in a resource pool being assigned to the host prior to the addition of roles.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":20
}
4.2.2. Enabling role(s)¶
To move a role, which is already added to a host, out of maintenance mode and thus enable it for use again, issue the following command:
# robin host enable-role [<host>] [<roles>]
|
Fully qualified hostname |
|
Comma separated list of roles to enable. Valid values are ‘storage’ and ‘compute’ |
Example:
# robin host enable-role centos-60-212.robinsystems.com Compute
Role(s) 'Compute' enabled on host centos-60-212.robinsystems.com
Enables a role that was previously disabled on a host.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: enable-role- This mandatory field within the payload specifies that the enable role operation is to be performed.roles: <list_of_roles>- This mandatory field within the payload is a list of roles that should be enabled on the specified host. Valid values include ‘compute’ and ‘storage’.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response: On success the reponse is empty.
4.2.3. Disabling role(s)¶
Within Robin when one disables a role, the role is said to be put into maintenance mode. This in turn means that for all intents and purposes the host does not have access to this role. This is useful for debugging purposes and to temporarily reserve the hosts resources. To move a role into maintenance mode and thus disable it for use, issue the following command:
# robin host disable-role [<host>] [<roles>]
|
Fully qualified hostname |
|
Comma separated list of roles to enable. Valid values are ‘storage’ and ‘compute’ |
Example:
# robin host disable-role centos-60-212.robinsystems.com Compute
Role(s) 'Compute' disabled on host centos-60-212.robinsystems.com
Disables a role for a host such that the host temporarily does not have access to it.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: disable-role- This mandatory field within the payload specifies that the disable role operation is to be performed.roles: <list_of_roles>- This mandatory field within the payload is a list of roles that should be disabled on the specified host. Valid values include ‘compute’ and ‘storage’.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response: On success the reponse is empty.
4.2.4. Removing role(s)¶
In order to a remove a role that had previously been assigned to a node, issue the following command:
# robin host remove-role [<hosts>] [<roles>]
--force
--yes
|
Comma separated list of hosts to remove role from. |
|
Comma separated list of roles. Valid values are ‘storage’, ‘compute’, and ‘manager’ |
|
Required if retrying a remove-role operation. |
|
Do not prompt the user for confirmation of removal |
Example:
# robin host remove-role centos-60-212 Compute,Storage --wait
Job: 192 Name: HostRemoveRoles State: VALIDATED Error: 0
Job: 192 Name: HostRemoveRoles State: WAITING Error: 0
Job: 192 Name: HostRemoveRoles State: COMPLETED Error: 0
Removes a role that had previously been assigned to a host.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: remove_roles- This mandatory field within the payload specifies that the remove role operation is to be performed.roles: <list_of_roles>- This mandatory field within the payload is a list of roles that should be removed from the specified host. Valid values include ‘storage’, ‘compute’, and ‘manager’.force: true- This field is mandatory when retrying the role removal operation, otherwise it can be excluded.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":24
}
4.3. Managing Resource Pools¶
Resource pools are a construct in Robin which allow you to group nodes in the cluster together for allocation purposes and consequently provide resource isolation. In turn this isolation has a multi-layered impact; primarily it determines what resources a tenant (and its associated users) has access to, which in turn affects the type of applications that can be deployed. As a result, resource pools can provide isolation at a user level as well as an application level since workloads can be manipulated such that they are deployed on a particular set of nodes. This is useful when one needs performant applications to run on more powerful or specifically configured nodes for example. The flip side of this is also true as nodes can be cordoned off from application deployments. The aforementioned isolation is maintained by the fact each node can be only be part of a single resource pool.
Upon installation, Robin creates a resource pool called default, which has no nodes associated with it and is attached to the Adminstrator’s tenant, for convienence sake.
Note
Nodes need to be added to resource pools before they can be initialized via roles and once they are added to a resource pool they cannot be removed.
The following commands are described in this section:
|
Add a resource pool |
|
List all resource pools |
|
Delete a resource pool |
|
Assign a resource pool to one or more hosts |
4.3.1. Adding a resource pool¶
In order to create an empty resource pool, issue the following the command:
# robin rpool add [<name>]
--zone <zone>
--tenant <tenant>
--desc <description>
|
Resource pool name |
|
Name of zone the resource pool belongs to |
|
Name of tenant the resource pool should be assigned to |
|
Description for resource pool |
Example:
# robin rpool add demo --desc "Resource pool for demo" --tenant Administrators
Added resource pool demo.
Creates an empty resource pool.
End Point: /api/v3/robin_server/rpools
Method: POST
URL Parameters: None
Data Parameters:
name: <rpool_name>- This mandatory field within the payload specifies the name of the resource pool to be created.description: <description>- This mandatory field within the payload specifies the description of the resource pool to be created.zone: <zone_name>- This mandatory field within the payload specifies the name of the zone in which the resource pool should be created.tenant: <tenant>- Utilizing this parameter within the payload, by specifying a tenant name, results in the consequent resource pool being associated with the aforementioned tenant.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error), 409 (Duplicate Resource Error)
Example Response:
Output
{
"message":"Added resource pool demo.\n"
}
4.3.2. Listing all resource pools¶
In order to list all resource pools currently present on the cluster along with details off their states, number of applications etc, issue the following command:
# robin rpool list [<name>]
--full
--json
|
Resource pool name to filter by. Note this is an optional argument |
|
Display additional information about each resource pool |
|
Output in JSON |
Example:
# robin rpool list
ID | Name | %cores | %mem | Nodes-ON/OF/FA | Apps-ON/OF/FA/AW/OT | Description
---+---------+--------+--------+----------------+---------------------+------------------------
1 | default | 20/100 | 20/100 | 3- 3/ 0/ 0 | 4- 4/ 0/ 0/ 0/ 0 | Default Resource Pool
2 | demo | 20/100 | 20/100 | 0- 0/ 0/ 0 | 0- 0/ 0/ 0/ 0/ 0 | Resource pool for demo
(ON: Online, OF: Offline, FA: Faulted, AW: Admin Wait, OT: Other states
Returns information on all resource pools currently present on the cluster along with details off their states, number of applications etc.
End Point: /api/v3/robin_server/rpools
Method: GET
URL Parameters: None
Data Parameters: None
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error)
Example Response:
Output
{
"items":[
{
"zoneid":1597147518,
"hdd_size":429496729600,
"apps":[
{
"name":"demo",
"state":6
}
],
"id":1,
"name":"default",
"ssd_size":0,
"num_cores":80,
"nodes":[
{
"zoneid":1597147518,
"k8s_node_status":"Ready",
"name":"cscale-82-140",
"remove_taint":true,
"state":"ONLINE",
"status":"Ready",
"hostname":"cscale-82-140.robinsystems.com",
"nodeid":1,
"roles":[
[
"MANAGER",
"ONLINE",
"READY"
],
[
"STORAGE",
"ONLINE",
"READY"
],
[
"COMPUTE",
"ONLINE",
"READY"
]
]
},
{
"zoneid":1597147518,
"k8s_node_status":"Notready",
"name":"cscale-82-139",
"remove_taint":false,
"state":"UNREACHABLE",
"status":"Notready",
"hostname":"cscale-82-139.robinsystems.com",
"nodeid":2,
"roles":[
[
"STORAGE",
"OFFLINE",
"NOTREADY"
],
[
"COMPUTE",
"OFFLINE",
"NOTREADY"
]
]
}
],
"hdd_used_size":73000000000.0,
"num_nodes":2,
"ssd_used_size":0,
"mem_size":67111419904,
"description":"Default Resource Pool"
}
]
}
4.3.3. Removing a resource pool¶
In order to remove a resource pool from Robin, issue the following command:
Note
Only empty resource pools which are not attached to any tenants can be removed.
# robin rpool remove [<name>]
--yes
|
Resource pool name to remove |
|
Do not prompt the user for confirmation of removal |
Example:
# robin rpool remove demo
Deleted resource pool 'demo'.
Removes a resource pool from Robin.
End Point: /api/v3/robin_server/rpools/<rpool_name>
Method: DELETE
URL Parameters: None
Data Parameters: None
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"message":"Deleted resource pool demo.\n"
}
4.3.4. Assigning a resource pool to a host¶
When one assigns a resource pool to a host, it results in the resources attached to the host contributing to the resource pool. These resources include storage drives, CPUs and memory. In turn any application created by a user in a tenant that is associated with the resource pool will have the ability to utilize the aforementioned resources. In order to assign a resource pool to a host, issue the following command:
# robin host assign-rpool [<hosts>] [<rpool>]
|
A single host or a comma separated list of hosts to assign a resource pool too |
|
Name of resource pool to assign |
Example:
# robin host assign-rpool centos-60-212.robinsystems.com demo --wait
Job: 22 Name: HostAddResourcePool State: VALIDATED Error: 0
Job: 22 Name: HostAddResourcePool State: WAITING Error: 0
Job: 22 Name: HostAddResourcePool State: COMPLETED Error: 0
Assigns a resource pool to a host, such that the resources such as storage drives, CPU etc. attached to the host contribute to the resource pool and are available for use by applications to be deployed within that resource pool.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: add_rpool- This mandatory field within the payload specifies the add resource pool action is to be performed.host_rpool: <dict_of_key_value_pairs>–hostname: <hostname>- This mandatory key value pairing within thehost_rpooldictionary indicates which host the operation is to be performed on.rpool: <rpool_name>- This mandatory key value pairing within thehost_rpooldictionary indicates which resource pool should be assigned to the aformentioned host.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":109
}
4.4. Managing File Collections¶
File collections are repositories in a Robin cluster where application bundles and images required for the provisioning of applications are stored. Robin File Server, one of the core Robin management services, is responsible for managing file collections and the files they contain. Storage for each file collection is allocated from Robin managed storage. The volumes allocated for each file collection can be configured with or without replication, however it is highly recommended that replicated storage be used. This will ensure that there will not be any data loss should one of the hosts providing the storage go down or one of storage disks fail.
The following commands are described in this section:
|
Create a file collection |
|
Delete a file collection |
|
List all file collections |
|
Disable a file collection |
|
Enable a file collection |
|
Move the contents of a file collection |
4.4.1. Creating a file collection¶
In order to create a file collection in order to store logs, images, bundles etc, issue the following the command:
# robin collection create <media_type> <rpool>
--collection_type <collection_type>
--storage_type <storage_type>
--replicas <replicas>
--size <size>
--force
|
The media type of storage drives to allocate from. Options include ‘HDD’ or ‘SSD’ |
|
Name of the resource pool to allocate from. |
``–collection_type <collection_type> `` |
Type of collection to create. Options include ‘LOG_COLLECTION’ or ‘FILE_COLLECTION’ |
|
Type of storage to use. Options include ‘STORMGR_VOLUME’, ‘LOCAL’, or ‘STORAGE_ARRAY’ |
|
Replication factor of the collection volume |
|
Size of allocated storage volume |
|
Override the requirement that the number of replicas must be set to 3 |
Example:
# robin collection create HDD default --replicas 1 --size 5G --force --wait
Job: 216 Name: CollectionAdd State: PROCESSED Error: 0
Job: 216 Name: CollectionAdd State: WAITING Error: 0
Job: 216 Name: CollectionAdd State: COMPLETED Error: 0
Creates a file collection in order to store logs, images, bundles etc.
End Point: /api/v3/robin_server/collections
Method: POST
URL Parameters: None
Data Parameters:
media: <media_type>- This mandatory field within the payload specifies the media type of storage drives to allocate from. Options include ‘HDD’ or ‘SSD’.rpool: <resource_pool>- This mandatory field within the payload specifies the resource pool name from which to allocate from.collection_type: <collection_type>- This mandatory field within the payload specifies the type of collection create. Options include ‘LOG_COLLECTION’ or ‘FILE_COLLECTION’.storage_type: <storage_type>- This mandatory field within the payload specifies the type of storage to use. Options include ‘STORMGR_VOLUME’, ‘LOCAL’, or ‘STORAGE_ARRAY’.replicas: <replicas>- This mandatory field within the payload specifies the replication factor of the collection volume. Valid values include: 1, 2, and 3.size: <size>- Utilizing this parameter, by specifying the size of the collection volume in bytes, results in a volume of the aforementioned size being created. The default size is 50GB.force: true- Utilizing this parameter allows one to override the requirment that the number of replicas for a collection value must be 3.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"plan":{
"collection_type":"FILE_COLLECTION",
"storage":[
{
"size":5368709120,
"rpool":"default",
"faultdomain":"host",
"replication":1,
"media":"HDD"
}
],
"storage_type":"STORMGR_VOLUME",
"authorization_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJ0ZW5hbnRfaWQiOjEsInVzZXJfaWQiOjMsImV4cCI6MTYwMDIyMjY5NH0.qsv3pBCXAfGnS1JrhvncQiubEqQcyPhV2fzGse8aA-A",
"hostname":"vnode-95-42.robinsystems.com",
"collection_id":1600138723311,
"force":true,
"collection_name":"file-collection-1600138723311"
},
"jobid":48
}
4.4.2. Deleting a file collection¶
In order to delete a file collection, issue the following the command:
# robin collection delete <collection_id>
--force
--yes
|
The media type of storage drives to allocate from. Options include HDD or SSD |
|
Delete a collection forcibly if the backing storage is down |
|
Do not prompt the user for confirmation of deletion |
Example:
# robin collection delete 1583820760654 --yes --wait
Job: 221 Name: CollectionDelete State: VALIDATED Error: 0
Job: 221 Name: CollectionDelete State: COMPLETED Error: 0
Deletes a file collection.
End Point: /api/v3/robin_server/collections
Method: DELETE
URL Parameters: None
Data Parameters:
collection_id: <collection_id>- This mandatory field within the payload specifies the ID of the collection that should be deleted.force: true- Utilizing this parameter results in the collection being forcibly removed if the backing storage is down.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":81,
"plan":{
"collection_id":1597897988633,
"authorization_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJ1c2VyX2lkIjozLCJ0ZW5hbnRfaWQiOjEsImV4cCI6MTU5Nzk2NzM1M30.Tcv6pAb66lC2k0VJpLj4oRT1elK99Qi7WBv_7Iid_zo"
}
}
4.4.3. Listing all file collections¶
In order to view all file collections available on the cluster alongside details such the host it is mounted on, its size, replication factor etc, issue the following command:
# robin collection list --name <name>
--full
--json
|
File collection name to filter by |
|
Display additional information about each file collection |
|
Output in JSON |
Example:
# robin collection list
Collection Id | Collection Type | Storage Type | Name | Hostname | Status | LastOpr | Used | Size (GB) | Replication Factor
--------------+-----------------+----------------+-------------------------------+--------------------------+--------+---------+-------+-----------+--------------------
1583356117884 | FILE_COLLECTION | STORMGR_VOLUME | file-collection-1583356117884 | vnode36.robinsystems.com | READY | Online | 1.3M | 10 | 1
1583820760654 | FILE_COLLECTION | STORMGR_VOLUME | file-collection-1583820760654 | vnode36.robinsystems.com | READY | Online | 2.2M | 5 | 1
Returns information on all file collections available on the cluster alongside details such the host it is mounted on, its size, replication factor etc.
End Point: /api/v3/robin_server/collections
Method: GET
URL Parameters:
name=<collection_name>: Utilizing this parameter results in only information for the file collection with the specified name being returned.
Data Parameters: None
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error)
Example Response:
Output
{
"collections":[
{
"notifications":{
},
"size":5368709120,
"id":1,
"used":"1.3M",
"name":"file-collection-1597122699552",
"pathname":"\/usr\/local\/robin\/collections\/file-collection-1597122699552",
"state":"Online",
"status":"READY",
"timestamp":"August 10, 2020 22:11:39",
"collection_type":"FILE_COLLECTION",
"replication":1,
"storage_type":"STORMGR_VOLUME",
"hostname":"cscale-82-140.robinsystems.com",
"collection_id":1597122699552
}
]
}
4.4.4. Disabling a file collection¶
In order to disable a file collection such that it temporarily cannot be used/accessed, issue the following command:
# robin collection offline <collection_id>
|
File collection to disable |
Example:
# robin collection offline 1583820760654 --wait
Job: 218 Name: CollectionOffline State: PROCESSED Error: 0
Job: 218 Name: CollectionOffline State: COMPLETED Error: 0
# robin collection list
Collection Id | Collection Type | Storage Type | Name | Hostname | Status | LastOpr | Size (GB) | Replication Factor
--------------+-----------------+----------------+-------------------------------+--------------------------+-----------+---------+-----------+--------------------
1583356117884 | FILE_COLLECTION | STORMGR_VOLUME | file-collection-1583356117884 | vnode36.robinsystems.com | READY | Online | 10 | 1
1583820760654 | FILE_COLLECTION | STORMGR_VOLUME | file-collection-1583820760654 | vnode36.robinsystems.com | NOT_READY | Offline | 5 | 1
Disable a file collection such that it temporarily cannot be used/accessed.
End Point: /api/v3/robin_server/collections
Method: PUT
URL Parameters: None
Data Parameters:
action: offline- This mandatory field within the payload specifies that the offline operation is to be performed.collection_id: <collection_id>- This mandatory field within the payload specifies which collection should be disabled.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"plan":{
"authorization_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJ0ZW5hbnRfaWQiOjEsInVzZXJfaWQiOjMsImV4cCI6MTU5Nzc2NjE1NH0._D_s9sj_6HoOQkqytCGqlgtL6aIsp1oym6BOwntM8iU",
"origin":1,
"collection_id":1597122699552
},
"jobid":2073
}
4.4.5. Enabling a file collection¶
In order to enable a file collection such that it is usable again by remounting it on the master node of the cluster, issue the following command:
# robin collection online <collection_id>
|
File collection to enable |
Example:
# robin collection online 1583820760654 --wait
Job: 219 Name: CollectionOnline State: PROCESSED Error: 0
Job: 219 Name: CollectionOnline State: COMPLETED Error: 0
# robin collection list
Collection Id | Collection Type | Storage Type | Name | Hostname | Status | LastOpr | Size (GB) | Replication Factor
--------------+-----------------+----------------+-------------------------------+--------------------------+--------+---------+-----------+--------------------
1583356117884 | FILE_COLLECTION | STORMGR_VOLUME | file-collection-1583356117884 | vnode36.robinsystems.com | READY | Online | 10 | 1
1583820760654 | FILE_COLLECTION | STORMGR_VOLUME | file-collection-1583820760654 | vnode36.robinsystems.com | READY | Online | 5 | 1
Enables a file collection such that it can be used again by remounting it on the master node of the cluster.
End Point: /api/v3/robin_server/collections
Method: PUT
URL Parameters: None
Data Parameters:
action: online- This mandatory field within the payload specifies that the online operation is to be performed.collection_id: <collection_id>- This mandatory field within the payload specifies which collection should be enabled.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":2108,
"plan":{
"origin":1,
"collection_id":1597122699552,
"authorization_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJ0ZW5hbnRfaWQiOjEsInVzZXJfaWQiOjMsImV4cCI6MTU5Nzc2NjE1NH0._D_s9sj_6HoOQkqytCGqlgtL6aIsp1oym6BOwntM8iU"
}
}
4.4.6. Moving a file collection¶
In order to move the resources contained within one file collection into another, issue the following command:
# robin collection move <collection_id> <target_collection_id>
|
ID of source file collection from which to move resources from |
|
ID of target file collection to move resource to |
Example:
# robin collection move 1583820760654 1583356117884 --wait
Job: 225 Name: CollectionMove State: PROCESSED Error: 0
Job: 225 Name: CollectionMove State: COMPLETED Error: 0
Moves the resources contained within one file collection into another.
End Point: /api/v3/robin_server/collections
Method: PUT
URL Parameters: None
Data Parameters:
action: moveresources- This mandatory field within the payload specifies that the resources within the source collection should be moved.collection_id: <collection_id>- This mandatory field within the payload specifies the source collection the resources should be moved from.target_collection_id: <target_collection_id>- This mandatory field within the payload specifies the destination collection the resources should be moved to.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":47,
"plan":{
"authorization_token":"eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJ0ZW5hbnRfaWQiOjEsInVzZXJfaWQiOjMsImV4cCI6MTYwNTI0NjA3OH0.wSV2mDr7y-kn_1MgafG7HxvjEdgu-cbTg3bSP4tQoZo",
"target_collection_id":1605164780025,
"origin":1,
"collection_id":1605164771262
}
}
4.5. Gathering information on nodes¶
Robin exposes multiple endpoints that provides a user the means by which to attain information about the hosts registered with a Robin cluster. The information that is returned is a combination of the physical attributes (obtained by the resource discovery described here), resource utilization and status of services for the host. This gives a user insight into the state of the cluster alongside granular details for each individual host and enables application deployment planning.
The following commands are described in this section:
|
View all hosts in a cluster |
|
Display detailed information about a host |
4.5.1. List all hosts¶
In order to view all hosts within a cluster alongside information on their statuses (from Robin’s perspective), resource consumption, and roles within the cluster, issue the following command:
# robin host list --services
--resources
--tags
--json
|
Show status information for each host |
|
Show resource utilization for each host |
|
Show tag information for each host |
|
Output in JSON |
Example 1 (Listing all hosts):
# robin host list
Id | Hostname | Version | Status | RPool | LastOpr | Roles | Isol Cores(SHR/DED/Total) | Non-Isol Cores | GPUs | Mem(Free/Alloc/Total) | HDD(#/Alloc/Total) | SSD(#/Alloc/Total) | Pod Usage | Joined Time
-------------+-------------------------------+-----------+--------+---------+---------+--------+---------------------------+----------------+------+-----------------------+--------------------+--------------------+-----------+----------------------
1596566663:1 | cscale-82-81.robinsystems.com | 5.3.0-172 | Ready | default | ONLINE | M*,S,C | 0/0/0 | 1/40 | 0/0 | 24G/6G/31G | 1/5G/100G | -/-/- | 11/110 | 04 Aug 2020 04:44:47
1596566663:2 | cscale-82-82.robinsystems.com | 5.3.0-172 | Ready | default | ONLINE | M,S,C | 0/0/0 | 2/40 | 0/0 | 24G/7G/31G | 2/40G/200G | -/-/- | 13/110 | 04 Aug 2020 04:51:52
1596566663:3 | cscale-82-83.robinsystems.com | 5.3.0-172 | Ready | default | ONLINE | M,S,C | 0/0/0 | 1/40 | 0/0 | 24G/6G/31G | 2/-/200G | -/-/- | 11/110 | 04 Aug 2020 04:58:25
1596566663:4 | qct-07.robinsystems.com | 5.3.0-172 | Ready | workers | ONLINE | S,C | 8/68/76 | 1/4 | 0/0 | 280G/95G/376G | 1/554G/893G | -/-/- | 83/110 | 04 Aug 2020 05:05:24
1596566663:5 | qct-08.robinsystems.com | 5.3.0-172 | Ready | workers | ONLINE | S,C | 0/76/76 | 1/4 | 0/0 | 284G/91G/376G | 1/547G/893G | -/-/- | 88/110 | 04 Aug 2020 05:12:06
1596566663:6 | qct-11.robinsystems.com | 5.3.0-172 | Ready | workers | ONLINE | S,C | 0/27/76 | 1/4 | 0/0 | 147G/40G/187G | 1/581G/893G | -/-/- | 34/110 | 04 Aug 2020 05:18:47
1596566663:7 | cscale-82-80.robinsystems.com | 5.3.0-172 | Ready | workers | ONLINE | C,S | 0/21/36 | 1/40 | 0/0 | 0.7G/30G/31G | 1/-/100G | -/-/- | 28/110 | 04 Aug 2020 20:21:40
Example 2 (Retrieving status information):
# robin host list --services
+-------------------------------+-------+-------+------+-------+-------+------+-------+------+------+------+-------+------+------+-------+---------+---------+
| Host | ConCl | ConSr | GCli | Httpd | Iomgr | RMon | Pgsql | RAgt | RAer | REvt | RFile | NMon | RSer | RWdog | Metrics | Stormgr |
+-------------------------------+-------+-------+------+-------+-------+------+-------+------+------+------+-------+------+------+-------+---------+---------+
| cscale-82-81.robinsystems.com | DOWN | UP | UP | UP | UP | UP | UP | UP | UP | UP | UP | UP | UP | UP | DOWN | UP |
| cscale-82-82.robinsystems.com | DOWN | UP | UP | UP | UP | UP | UP | UP | DOWN | DOWN | DOWN | DOWN | DOWN | UP | DOWN | DOWN |
| cscale-82-83.robinsystems.com | DOWN | UP | UP | UP | UP | UP | UP | UP | DOWN | DOWN | DOWN | DOWN | DOWN | UP | DOWN | DOWN |
| qct-07.robinsystems.com | UP | DOWN | UP | UP | UP | UP | DOWN | UP | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN |
| qct-08.robinsystems.com | UP | DOWN | UP | UP | UP | UP | DOWN | UP | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN |
| qct-11.robinsystems.com | UP | DOWN | UP | UP | UP | UP | DOWN | UP | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN |
| cscale-82-80.robinsystems.com | UP | DOWN | UP | UP | UP | UP | DOWN | UP | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN | DOWN |
+-------------------------------+-------+-------+------+-------+-------+------+-------+------+------+------+-------+------+------+-------+---------+---------+
UP: Running
CRIT: Critical and Down
DOWN: Not Running
Returns information on all hosts within a cluster including details on their statuses (from Robin’s perspective), resource consumption, and roles within the cluster.
End Point: /api/v5/robin_server/hosts
Method: GET
URL Parameters:
details=tags: Utilizing this parameter results in tag information for each host being present in the response payload.
Data Parameters: None
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error)
Example Response:
Output
{
"items":[
{
"memory_used":2692743168,
"memory":33555709952,
"isol_shared_map":{
},
"zoneid":1596601846,
"non_isol_cores_used":3,
"pods_used":26,
"rack":"default",
"napps":2,
"non_isol_total":400,
"k8s_node_name":"cscale-82-140",
"mem_for_storage":1073741824,
"id":1,
"lab":"default",
"gpu_cores_allocated":0,
"isol_dedicated_cores_used":0,
"roles":[
[
"MANAGER",
"ONLINE",
"READY"
],
[
"COMPUTE",
"ONLINE",
"READY"
],
[
"STORAGE",
"ONLINE",
"READY"
]
],
"cpu_cores_used":0,
"cpu_prov_factor":10,
"services":"{\"update_time\":1596761892.3700919151,\"services\":{\"consul_dns\":true,\"stormgr-server\":{\"Id\":\"stormgr-server\",\"MainPID\":2299,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:31:26.794916\",\"ActiveState\":\"active\"},\"gui-cli\":{\"Id\":\"gui-cli\",\"MainPID\":2647,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:31:32.291459\",\"ActiveState\":\"active\"},\"consul-client\":{\"Id\":\"consul-client\",\"MainPID\":0,\"Type\":\"simple\",\"ExecMainStartTimestamp\":0,\"ActiveState\":\"inactive\"},\"httpd\":{\"Id\":\"httpd\",\"MainPID\":2613,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:31:32.124472\",\"ActiveState\":\"active\"},\"robin-node-monitor\":{\"Id\":\"robin-node-monitor\",\"MainPID\":1278,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:31:13.459063\",\"ActiveState\":\"active\"},\"iomgr-server\":{\"Id\":\"iomgr-server\",\"MainPID\":7384,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:55:13.851492\",\"ActiveState\":\"active\"},\"robin-event-server\":{\"Id\":\"robin-event-server\",\"MainPID\":1039,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:30:55.322709\",\"ActiveState\":\"active\"},\"consul-server\":{\"Id\":\"consul-server\",\"MainPID\":564,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:30:32.432940\",\"ActiveState\":\"active\"},\"consul_members\":[{\"DelegateMax\":5,\"ProtocolMin\":1,\"Port\":29460,\"Status\":1,\"ProtocolMax\":5,\"DelegateCur\":4,\"ProtocolCur\":2,\"Name\":\"cscale-82-140.robinsystems.com\",\"Tags\":{\"dc\":\"consul\",\"role\":\"consul\",\"vsn\":\"2\",\"wan_join_port\":\"29461\",\"segment\":\"\",\"port\":\"29459\",\"raft_vsn\":\"2\",\"vsn_min\":\"2\",\"vsn_max\":\"3\",\"id\":\"9dbc13cd-bbb4-1bf1-9bcd-f3d7e0f0026f\",\"bootstrap\":\"1\",\"build\":\"0.9.4:40f243a+\"},\"Addr\":\"10.9.82.140\",\"DelegateMin\":2}],\"robin-file-server\":{\"Id\":\"robin-file-server\",\"MainPID\":1071,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:30:55.801664\",\"ActiveState\":\"active\"},\"robin-watchdog\":{\"Id\":\"robin-watchdog\",\"MainPID\":860,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:30:37.982382\",\"ActiveState\":\"active\"},\"sherlock-server\":{\"Id\":\"sherlock-server\",\"MainPID\":0,\"Type\":\"simple\",\"ExecMainStartTimestamp\":0,\"ActiveState\":\"inactive\"},\"robin-agent\":{\"Id\":\"robin-agent\",\"MainPID\":9186,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:59:23.165676\",\"ActiveState\":\"active\"},\"postgresql-9.6\":{\"Id\":\"postgresql-9.6\",\"MainPID\":660,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:30:34.988682\",\"ActiveState\":\"active\"},\"monitor-server\":{\"Id\":\"monitor-server\",\"MainPID\":2687,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:31:32.455445\",\"ActiveState\":\"active\"},\"robin-server\":{\"Id\":\"robin-server\",\"MainPID\":64400,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-06 17:57:55.208502\",\"ActiveState\":\"active\"},\"robin-auth-server\":{\"Id\":\"robin-auth-server\",\"MainPID\":1010,\"Type\":\"simple\",\"ExecMainStartTimestamp\":\"2020-08-04 14:30:50.844131\",\"ActiveState\":\"active\"}}}",
"gpu_cores":0,
"sysmem":[
33555709952,
2689277952,
7223099392,
0,
0,
0,
22075457536,
843776
],
"ssd_faulted":0,
"isol_shared_cores_used":0,
"hugepages_1g":0,
"ninstances":2,
"maintenance_mode":"DISABLED",
"memory_allocated":0,
"hostname":"cscale-82-140.robinsystems.com",
"datacenter":"default",
"tags":{
"kubernetes.io\/os":[
"linux"
],
"robin.io\/robinrpool":[
"default"
],
"kubernetes.io\/arch":[
"amd64"
]
},
"remove_taint":true,
"hugepages_2m_allocated":0,
"rpool":"default",
"pods":110,
"state":"ONLINE",
"name":"cscale-82-140",
"rcm_ha_role":"MANAGER_MASTER",
"isol_total":0,
"hugepages_2m":0,
"hdd_faulted":0,
"ipaddresses":[
{
"mac_address":"00:15:5d:14:06:0e",
"netmask":"255.255.0.0",
"ip_address":"10.9.82.140"
}
],
"cpu_cores_allocated":0,
"memory_reserved":6442450944.0,
"cpu_cores":40,
"hugepages_1g_allocated":0,
"k8s_node_status":"Ready",
"status":"Ready",
"nics":[
{
"allowed_vlans":[
],
"function":null,
"numa_node":null,
"vendor":null,
"mtu":1500,
"mac_address":"00:15:5d:14:06:0e",
"bus":null,
"name":"br0",
"physical_nic":"eth0",
"num_vfs":0,
"linkstate":"",
"all_vlans_allowed":false,
"used_vfs":0,
"native_vfdriver":null,
"native_vlan":null,
"vendor_desc":null,
"domain":null,
"untagged":false,
"slot":null,
"vfdrivers":[
]
}
],
"public_hostname":"cscale-82-140.robinsystems.com",
"cpu_cores_present":40,
"sysinfo":{
"join_time":1596576678,
"current_version":"5.3.0-171",
"iqn":"iqn.1994-05.com.redhat:329b8568de1",
"install_date":"Tue Mar 17 23:49:17 UTC 2020",
"wwpns":[
],
"distribution":"CentOS Linux",
"version":"#1 SMP Tue Mar 17 23:49:17 UTC 2020",
"uuid":"",
"boot_time":1596576222,
"robin_software":[
{
"version":"5.3.0",
"patch":"",
"full_version":"5.3.0-171",
"install_date":"2020-08-03",
"patch_date":"",
"release":"171",
"build_info":"robin-c2edf85eaa83a42ced9512e7de9c7c2f1e4fa962:robin-ui:9ee33fd00273ba19861d4dc3ef8c6169d822d3e0:robingraph:cf0ceefe696ccac2dbd2eeb1d28b859955452843"
}
],
"release":"3.10.0-1062.18.1.el7.x86_64",
"system":"Linux",
"processor":"x86_64"
},
"disks":[
{
"spf":0.8,
"state":"READY",
"type":"HDD",
"zoneid":1596601846,
"dev":"\/dev\/sdb",
"max_alloc_slices":77,
"free_alloc_slices":68,
"model":null,
"allocated":7,
"maintenance_mode":"DISABLED",
"max_throughput_intensive_vols_per_disk":1,
"role":"Storage",
"wwn":"0x600224804c48fd7e16c608dea0919064",
"status":"ONLINE",
"make":null,
"devpath":"\/dev\/disk\/by-id\/scsi-3600224804c48fd7e16c608dea0919064",
"alloc_slices":9,
"reattachable":0,
"max_volumes_per_disk":10,
"protected":0,
"capacity":107374182400,
"node_ref":1,
"max_latency_sensitive_vols_per_disk":2,
"pused":234881024,
"pfree":104287174656
},
{
"spf":0.8,
"state":"READY",
"type":"HDD",
"zoneid":1596601846,
"dev":"\/dev\/sdc",
"max_alloc_slices":77,
"free_alloc_slices":53,
"model":null,
"allocated":20,
"maintenance_mode":"DISABLED",
"max_throughput_intensive_vols_per_disk":1,
"role":"Storage",
"wwn":"0x600224803bcdafde95b1f5cd27ceb5fb",
"status":"ONLINE",
"make":null,
"devpath":"\/dev\/disk\/by-id\/scsi-3600224803bcdafde95b1f5cd27ceb5fb",
"alloc_slices":24,
"reattachable":0,
"max_volumes_per_disk":10,
"protected":0,
"capacity":107374182400,
"node_ref":1,
"max_latency_sensitive_vols_per_disk":2,
"pused":939524096,
"pfree":103582531584
},
{
"spf":0.8,
"state":"INIT",
"type":"HDD",
"zoneid":1596601846,
"dev":"\/dev\/dm-1",
"max_alloc_slices":5,
"free_alloc_slices":5,
"model":null,
"allocated":0,
"maintenance_mode":"DISABLED",
"max_throughput_intensive_vols_per_disk":1,
"role":"RootDisk",
"wwn":"0x600224801d3ac9b6650afd3280aa5898-centos-swap",
"status":"UNKNOWN",
"make":null,
"devpath":"\/dev\/disk\/by-id\/dm-uuid-LVM-vI83PDTxV3H0dWyAXfH5ef7rxTOuYyphaFy4aq3EUo1yluonS8FG0LF16ycBrdEw",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"protected":0,
"capacity":8254390272,
"node_ref":1,
"max_latency_sensitive_vols_per_disk":2,
"pused":0,
"pfree":0
},
{
"spf":0.8,
"state":"INIT",
"type":"HDD",
"zoneid":1596601846,
"dev":"\/dev\/dm-0",
"max_alloc_slices":38,
"free_alloc_slices":38,
"model":null,
"allocated":0,
"maintenance_mode":"DISABLED",
"max_throughput_intensive_vols_per_disk":1,
"role":"RootDisk",
"wwn":"0x600224801d3ac9b6650afd3280aa5898-centos-root",
"status":"UNKNOWN",
"make":null,
"devpath":"\/dev\/disk\/by-id\/dm-uuid-LVM-vI83PDTxV3H0dWyAXfH5ef7rxTOuYyphgpZcvqGdfOKaXbEbOZzNthc6btsoSXDj",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"protected":0,
"capacity":53687091200,
"node_ref":1,
"max_latency_sensitive_vols_per_disk":2,
"pused":0,
"pfree":0
},
{
"spf":0.8,
"state":"INIT",
"type":"HDD",
"zoneid":1596601846,
"dev":"\/dev\/dm-2",
"max_alloc_slices":32,
"free_alloc_slices":32,
"model":null,
"allocated":0,
"maintenance_mode":"DISABLED",
"max_throughput_intensive_vols_per_disk":1,
"role":"RootDisk",
"wwn":"0x600224801d3ac9b6650afd3280aa5898-centos-home",
"status":"UNKNOWN",
"make":null,
"devpath":"\/dev\/disk\/by-id\/dm-uuid-LVM-vI83PDTxV3H0dWyAXfH5ef7rxTOuYyphQObDlS6eMUSpSxH5zsvyg9I5a0Gpuj5W",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"protected":0,
"capacity":44350570496,
"node_ref":1,
"max_latency_sensitive_vols_per_disk":2,
"pused":0,
"pfree":0
},
{
"spf":0.8,
"state":"INIT",
"type":"HDD",
"zoneid":1596601846,
"dev":"\/dev\/sda",
"max_alloc_slices":77,
"free_alloc_slices":77,
"model":null,
"allocated":0,
"maintenance_mode":"DISABLED",
"max_throughput_intensive_vols_per_disk":1,
"role":"RootDisk",
"wwn":"0x600224801d3ac9b6650afd3280aa5898",
"status":"UNKNOWN",
"make":null,
"devpath":"\/dev\/disk\/by-id\/scsi-3600224801d3ac9b6650afd3280aa5898",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"protected":0,
"capacity":107374182400,
"node_ref":1,
"max_latency_sensitive_vols_per_disk":2,
"pused":0,
"pfree":0
}
]
}
],
"total":1,
"page_num":1,
"nodes_count":1,
"num_items":1,
"page_size":1
}
4.5.2. Show information about a specific host¶
In order to display detailed information for a host such as the storage allocation breakdown, discovered physical attributes with their utilization (NUMA configuration, network topology etc.) and service details, issue the following command:
# robin host info <hostname>
--services
--resources
--config
--consul
--json
|
FQDN of host |
|
Show status information for the host |
|
Show resource utilization for the host |
|
Show config info |
|
Show consul cluster info |
|
Output in JSON |
Example:
# robin host info poch01.robin.io
Output
Host: qct-07.robinsystems.com
Zone Id: 1596566663
Host Id: 4
Type: physical
Version: 5.3.0-172
Kernel Version: 3.10.0-1062.el7.x86_64
Boot Time: 04 Aug 2020 03:18:01
Resource pool: workers
CPU:
Total Cores: 80
Total Isolated Cores: 76
Total Non-Isolated Cores: 4
Non-Isolated CPUs allocated: 1
Shared Isolated CPUs allocated: 8
Dedicated Isolated CPUs allocated: 68
Provisioning Factor: 1
NUMA Topology:
Node 0:
Total Memory: 187G
Total Isolated CPUs: 38
Total Non-Isolated CPUs: 2
Total Reserved CPUs: 0
Non-Isolated Pinned CPUs: 0
Isolated Shared Pinned CPUs: 0
Isolated Dedicated Pinned CPUs: 38
Total HugePages_1G: -
Total HugePages_2M: -
CPU List: 1-19,41-59
NIC List: enp94s0f0,enp59s0f0,enp59s0f1,enp94s0f1
Node 1:
Total Memory: 188G
Total Isolated CPUs: 38
Total Non-Isolated CPUs: 2
Total Reserved CPUs: 0
Non-Isolated Pinned CPUs: 0
Isolated Shared Pinned CPUs: 8
Isolated Dedicated Pinned CPUs: 30
Total HugePages_1G: -
Total HugePages_2M: -
CPU List: 21-39,61-79
NIC List: enp175s0f1,enp175s0f0
GPU:
Total Cores: 0
Memory:
System Total: 376G
Allocatable Total: 376G
Reserved: 6G
Robin Manager services: -
Robin Compute services: 4G
Robin Storage services: 2G
Memory allocated to instances: 88G
Free Total: 292G
HugePages_2M:
Total: -
Allocated for Robin apps: -
HugePages_1G:
Total: -
Allocated for Robin apps: -
POD Utilization: 83/110
Network:
Bridge Interface: br0
Physical Interface: enp94s0f1
MTU: 1500
Product Info: 158B - Ethernet Controller XXV710 for 25GbE SFP28 (Ethernet Network Adapter XXV710)
Vendor Info: 8086 - Intel Corporation
NUMA Node: 0
H/W Info: 0000:5e:00.1
IP Addresses: 10.9.20.15/16
Interface: enp175s0f1
MTU: 1500
Product Info: 158B - Ethernet Controller XXV710 for 25GbE SFP28 (Ethernet Network Adapter XXV710)
Vendor Info: 8086 - Intel Corporation
NUMA Node: 1
H/W Info: 0000:af:00.1
Interface: enp175s0f0
MTU: 1500
Product Info: 158B - Ethernet Controller XXV710 for 25GbE SFP28 (Ethernet Network Adapter XXV710-2)
Vendor Info: 8086 - Intel Corporation
NUMA Node: 1
H/W Info: 0000:af:00.0
Interface: enp94s0f0
MTU: 1500
Product Info: 158B - Ethernet Controller XXV710 for 25GbE SFP28 (Ethernet Network Adapter OCP XXV710-2)
Vendor Info: 8086 - Intel Corporation
NUMA Node: 0
H/W Info: 0000:5e:00.0
Interface: enp59s0f0
MTU: 1500
Product Info: 158B - Ethernet Controller XXV710 for 25GbE SFP28 (Ethernet Network Adapter XXV710-2)
Vendor Info: 8086 - Intel Corporation
NUMA Node: 0
H/W Info: 0000:3b:00.0
Interface: enp59s0f1
MTU: 1500
Product Info: 158B - Ethernet Controller XXV710 for 25GbE SFP28 (Ethernet Network Adapter XXV710)
Vendor Info: 8086 - Intel Corporation
NUMA Node: 0
H/W Info: 0000:3b:00.1
Public IP Address: 10.9.20.15
Public Hostname: qct-07.robinsystems.com
Instances: 69
State: ONLINE
Status: Ready
K8s_Node_Status: Ready
Maintenance Mode: DISABLED
Consul state: UP
Roles:
STORAGE: ONLINE Status:READY
COMPUTE: ONLINE Status:READY
Storage:
Type | Used (GB) | Robin Allocated (GB) | K8s Allocated (GB) | Total (GB)
-----+-----------+----------------------+--------------------+------------
HDD | 38 | 538 | 16 | 893
SSD | - | - | - | -
Services:
Name | State | RoleTags | PID | Started
-------------------+-------+----------+------+----------------------------
consul-client | UP | A | 351 | 2020-08-04 17:51:31.538657
consul-server | DOWN | M | 0 | 0
gui-cli | UP | - | 1132 | 2020-08-04 17:51:50.479143
httpd | UP | * | 1077 | 2020-08-04 17:51:50.116133
iomgr-server | UP | C S | 5391 | 2020-08-04 17:52:16.467809
monitor-server | UP | M C S | 539 | 2020-08-04 17:51:42.748944
postgresql-9.6 | DOWN | M | 0 | 0
robin-agent | UP | M C S | 600 | 2020-08-04 17:51:45.394012
robin-auth-server | DOWN | M* | 0 | 0
robin-event-server | DOWN | M* | 0 | 0
robin-file-server | DOWN | M* | 0 | 0
robin-node-monitor | DOWN | M* | 0 | 0
robin-server | DOWN | M* | 0 | 0
robin-watchdog | DOWN | M | 0 | 0
sherlock-server | DOWN | - | 0 | 0
stormgr-server | DOWN | M* | 0 | 0
Last updated (04 Aug 2020 18:40:14)
UP: Running
CRIT: Critical and Down
DOWN: Not Running
Root Disk storage info:
Partition | Name | Size (GB) | Available (GB)
--------------------------+-----------------+-----------+----------------
/var/log | RobinLog | 299 | 268
/var/lib/pgsql | Pgsql | 299 | 268
/var/crash | Crash | 299 | 268
/var/lib/robin | RobinLib | 299 | 268
/var/lib/[appropriateCRI] | ContainerImages | 15 | -
Unused container images: 6G
Image | Size (GB)
--------------------------------------------+-----------
k8s.gcr.io/kube-controller-manager:v1.18.6 | 0.15
k8s.gcr.io/kube-apiserver:v1.18.6 | 0.16
k8s.gcr.io/kube-scheduler:v1.18.6 | 0.09
robinsys/robinimg:5.2.7-18 | 3
quay.io/k8scsi/csi-provisioner:v1.6.0_robin | 0.04
k8s.gcr.io/kube-proxy:v1.17.5 | 0.11
k8s.gcr.io/kube-apiserver:v1.17.5 | 0.16
k8s.gcr.io/kube-controller-manager:v1.17.5 | 0.15
k8s.gcr.io/kube-scheduler:v1.17.5 | 0.09
quay.io/k8scsi/snapshot-controller:v2.1.0 | 0.04
k8s.gcr.io/pause:3.2 | 0.0
prom/prometheus:v2.16.0 | 0.12
quay.io/k8scsi/csi-attacher:v2.1.0 | 0.04
calico/typha:v3.11.1 | 0.05
calico/pod2daemon-flexvol:v3.11.1 | 0.1
calico/cni:v3.11.1 | 0.18
calico/kube-controllers:v3.11.1 | 0.05
quay.io/k8scsi/csi-provisioner:v1.4.0_robin | 0.05
k8s.gcr.io/kube-proxy:v1.16.3 | 0.08
k8s.gcr.io/kube-apiserver:v1.16.3 | 0.2
k8s.gcr.io/kube-controller-manager:v1.16.3 | 0.15
k8s.gcr.io/kube-scheduler:v1.16.3 | 0.08
k8s.gcr.io/coredns:1.6.5 | 0.04
metallb/controller:v0.8.2 | 0.04
metallb/speaker:v0.8.2 | 0.04
k8s.gcr.io/etcd:3.4.3-0 | 0.27
quay.io/k8scsi/csi-snapshotter:v1.2.2 | 0.04
k8s.gcr.io/etcd:3.3.15-0 | 0.23
quay.io/k8scsi/csi-attacher:v1.2.1 | 0.04
k8s.gcr.io/coredns:1.6.2 | 0.04
robinsys/genie-plugin:v3.0 | 0.02
quay.io/k8scsi/csi-provisioner:v1.0.0_robin | 0.04
quay.io/k8scsi/csi-provisioner:v0.4.1_robin | 0.04
robinsys/coredns:1.2.2 | 0.03
Returns detailed information for a host such as the storage allocation breakdown, discovered physical attributes with their utilization (NUMA configuration, network topology etc.) and service details.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: GET
URL Parameters:
diskinfo=true: Utilizing this parameter results in details of the disks attached to the specified host being returned.
Data Parameters: None
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error)
Example Response:
Output
{
"items":[
{
"memory_used":2692743168,
"hdd_lalloc":35433480192.0,
"memory":33555709952,
"lab":"default",
"saas_mode":false,
"zoneid":1596601846,
"non_isol_total_with_prov":400,
"zone_name":"default",
"rack":"default",
"ipaddresses":[
{
"ip_address":"10.9.82.140",
"mac_address":"00:15:5d:14:06:0e",
"netmask":"255.255.0.0"
}
],
"public_ip":"10.9.82.140",
"ssd_nonrobin_usage":0,
"k8s_node_name":"cscale-82-140",
"mem_for_storage":1073741824,
"id":1,
"ssd_for_storage":0,
"rcm_ha_role":"MANAGER_MASTER",
"ssd_robin_usage":0,
"gpu_cores_allocated":0,
"numa_map":{
"0":{
"memory_used":0,
"hugepages_1g_used":0,
"isol_total":0,
"isol_shared_map":{
},
"cpu_reserved":0,
"numa_id":0,
"non_isol_cores_used":2,
"cpu_ids":"",
"cpu_used":0,
"mem_used":1493172224,
"non_isol_total":20,
"hugepages_2m_used":0,
"gpu_used":0,
"isol_shared_cores_used":0,
"hugepages_1g_total":0,
"cpu_total":20,
"hugepages_2m_total":0,
"memory_total":16777626965,
"isol_dedicated_cores_used":0
},
"1":{
"memory_used":0,
"hugepages_1g_used":0,
"isol_total":0,
"isol_shared_map":{
},
"cpu_reserved":0,
"numa_id":1,
"non_isol_cores_used":0,
"cpu_ids":"",
"cpu_used":0,
"mem_used":0,
"non_isol_total":20,
"hugepages_2m_used":0,
"gpu_used":0,
"isol_shared_cores_used":0,
"hugepages_1g_total":0,
"cpu_total":20,
"hugepages_2m_total":0,
"memory_total":16778082988,
"isol_dedicated_cores_used":0
}
},
"roles":[
[
"MANAGER",
"ONLINE",
"READY"
],
[
"COMPUTE",
"ONLINE",
"READY"
],
[
"STORAGE",
"ONLINE",
"READY"
]
],
"ssd_max_alloc_slices":0,
"cpu_prov_factor":10,
"services":{
"update_time":1596761892.3700919151,
"services":{
"consul_dns":true,
"stormgr-server":{
"MainPID":2299,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:31:26.794916",
"RoleTags":[
"M*"
],
"Id":"stormgr-server",
"State":"UP",
"ActiveState":"active"
},
"gui-cli":{
"MainPID":2647,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:31:32.291459",
"RoleTags":[
"-"
],
"Id":"gui-cli",
"State":"UP",
"ActiveState":"active"
},
"consul-client":{
"MainPID":0,
"Type":"simple",
"ExecMainStartTimestamp":0,
"RoleTags":[
"A"
],
"Id":"consul-client",
"State":"DOWN",
"ActiveState":"inactive"
},
"robin-server":{
"MainPID":64400,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-06 17:57:55.208502",
"RoleTags":[
"M*"
],
"Id":"robin-server",
"State":"UP",
"ActiveState":"active"
},
"robin-node-monitor":{
"MainPID":1278,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:31:13.459063",
"RoleTags":[
"M*"
],
"Id":"robin-node-monitor",
"State":"UP",
"ActiveState":"active"
},
"iomgr-server":{
"MainPID":7384,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:55:13.851492",
"RoleTags":[
"C",
"S"
],
"Id":"iomgr-server",
"State":"UP",
"ActiveState":"active"
},
"consul_members":[
{
"DelegateMax":5,
"ProtocolCur":2,
"Port":29460,
"Status":1,
"ProtocolMax":5,
"DelegateCur":4,
"Tags":{
"dc":"consul",
"role":"consul",
"vsn":"2",
"wan_join_port":"29461",
"segment":"",
"port":"29459",
"raft_vsn":"2",
"vsn_min":"2",
"vsn_max":"3",
"id":"9dbc13cd-bbb4-1bf1-9bcd-f3d7e0f0026f",
"bootstrap":"1",
"build":"0.9.4:40f243a+"
},
"ProtocolMin":1,
"Name":"cscale-82-140.robinsystems.com",
"Addr":"10.9.82.140",
"DelegateMin":2
}
],
"robin-file-server":{
"MainPID":1071,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:30:55.801664",
"RoleTags":[
"M*"
],
"Id":"robin-file-server",
"State":"UP",
"ActiveState":"active"
},
"robin-event-server":{
"MainPID":1039,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:30:55.322709",
"RoleTags":[
"M*"
],
"Id":"robin-event-server",
"State":"UP",
"ActiveState":"active"
},
"sherlock-server":{
"MainPID":0,
"Type":"simple",
"ExecMainStartTimestamp":0,
"RoleTags":[
"-"
],
"Id":"sherlock-server",
"State":"DOWN",
"ActiveState":"inactive"
},
"robin-agent":{
"MainPID":9186,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:59:23.165676",
"RoleTags":[
"M",
"C",
"S"
],
"Id":"robin-agent",
"State":"UP",
"ActiveState":"active"
},
"postgresql-9.6":{
"MainPID":660,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:30:34.988682",
"RoleTags":[
"M"
],
"Id":"postgresql-9.6",
"State":"UP",
"ActiveState":"active"
},
"consul-server":{
"MainPID":564,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:30:32.432940",
"RoleTags":[
"M"
],
"Id":"consul-server",
"State":"UP",
"ActiveState":"active"
},
"httpd":{
"MainPID":2613,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:31:32.124472",
"RoleTags":[
"*"
],
"Id":"httpd",
"State":"UP",
"ActiveState":"active"
},
"robin-auth-server":{
"MainPID":1010,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:30:50.844131",
"RoleTags":[
"M*"
],
"Id":"robin-auth-server",
"State":"UP",
"ActiveState":"active"
},
"robin-watchdog":{
"MainPID":860,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:30:37.982382",
"RoleTags":[
"M"
],
"Id":"robin-watchdog",
"State":"UP",
"ActiveState":"active"
},
"monitor-server":{
"MainPID":2687,
"Type":"simple",
"ExecMainStartTimestamp":"2020-08-04 14:31:32.455445",
"RoleTags":[
"M",
"C",
"S"
],
"Id":"monitor-server",
"State":"UP",
"ActiveState":"active"
}
}
},
"non_isol_total":40,
"gpu_cores":0,
"hdd_robin_usage":35433480192,
"visibledisks":[
"0x600224801d3ac9b6650afd3280aa5898",
"0x600224801d3ac9b6650afd3280aa5898-centos-root",
"0x600224801d3ac9b6650afd3280aa5898-centos-swap",
"0x600224801d3ac9b6650afd3280aa5898-centos-home",
"0x600224804c48fd7e16c608dea0919064",
"0x600224803bcdafde95b1f5cd27ceb5fb"
],
"ssd_faulted":0,
"isol_shared_cores_used":0,
"nic_details":{
"br0":{
"vfdrivers":[
],
"product_desc":null,
"all_vlans_allowed":false,
"mtu":1500,
"bus":null,
"vendor_id":null,
"slot":null,
"physical_nic":"eth0",
"allowed_vlans":[
],
"num_vfs":0,
"function":null,
"ips":[
"10.9.82.140\/16"
],
"product_id":null,
"local_cpulist":null,
"native_vlan":null,
"vendor_desc":null,
"domain":null,
"untagged":false,
"used_vfs":0,
"numa_node":null
}
},
"ninstances":2,
"non_isol_cores_used":3,
"maintenance_mode":"DISABLED",
"memory_allocated":0,
"mem_for_management":1073741824.0,
"rpool_id":1,
"hdd_for_storage":214748364800,
"datacenter":"default",
"tags":{
"kubernetes.io\/os":[
"linux"
],
"robin.io\/robinrpool":[
"default"
],
"kubernetes.io\/arch":[
"amd64"
]
},
"hugepages_1g":0,
"hugepages_2m_allocated":0,
"rpool":"default",
"hdd_total":428414599168,
"ssd_free_alloc_slices":0,
"pods":110,
"state":"ONLINE",
"status":"Ready",
"instances":[
{
"state":"STARTED",
"name":"rohan-app.nginx.03",
"hostname":"rohan-app-nginx-03.t001-u000003.svc.cluster.local"
},
{
"state":"STARTED",
"name":"test-RIC-1.server.01",
"hostname":"test-ric-1-server-01.t001-u000003.svc.cluster.local"
}
],
"isol_dedicated_cores_used":0,
"host_type":"physical",
"ssd_pused":0,
"isol_total":0,
"primary_ip":"10.9.82.140",
"isol_shared_map":{
},
"hugepages_2m":0,
"pods_used":26,
"hdd_lused":0,
"hdd_faulted":0,
"napps":2,
"hdd_nonrobin_usage":0,
"cpu_cores_present":40,
"ssd_total":0,
"cpu_cores_allocated":0,
"is_master":true,
"memory_reserved":6442450944.0,
"cpu_cores":400,
"config":{
"stormgr_rest_port":29454,
"monitor_host_mem_lowmark":0.8,
"monitor_host_root_volume_highmark":0.9,
"rio_rest_port":29456,
"stormgr_rest_listen_addr":"127.0.0.1",
"kvm_enabled":true,
"hard_reset_on_isolation":0,
"monitor_host_cpu_lowmark":0.8,
"monitor_host_var_crash_volume_highmark":0.9,
"monitor_interval":1,
"monitor_host_var_pgsql_volume_lowmark":0.5,
"kubelet_restart_bursttime":25,
"server_rest_port":29442,
"kubelet_restart_burstlimit":2,
"event_server_port":29449,
"rio_rpc_port":29453,
"rdvm_bmapcache_skip_all":0,
"rdvm_mem_maxcap":25769803776,
"rest_server":"cscale-82-140.robinsystems.com",
"registration_timeout":10,
"rdvm_rpc_port":29452,
"node_exporter_port":29457,
"stormgr_rpc_port":29451,
"monitor_host_root_volume_lowmark":0.85,
"database_port":29458,
"rdvm_rest_listen_addr":"127.0.0.1",
"https_port":29443,
"metrics_grafana_details":"{\"url\": \"\", \"auth\": \":\"}",
"monitor_host_var_volume_lowmark":0.85,
"monitor_host_cpu_highmark":0.85,
"rio_rest_listen_addr":"127.0.0.1",
"monitor_host_var_robin_volume_highmark":0.9,
"rdvm_mem_alloc":1073741824,
"monitor_num_samples":3600,
"monitor_host_swap_lowmark":0.75,
"watchdog_loop_interval":3,
"rdvm_rest_port":29455,
"monitor_container_swap_highmark":0.8,
"consul_serfwan_port":29461,
"saas_mode":false,
"file_object_cache":"\/var\/lib\/robin\/file_object_cache",
"node_monitor_port":29467,
"monitor_influx_details":"{\"url\": \"\", \"dbname\": \"robin\", \"auth\": \":\" }",
"consul_http_port":29462,
"monitor_container_swap_lowmark":0.75,
"hostname":"cscale-82-140.robinsystems.com",
"monitor_host_var_volume_highmark":0.9,
"network_type":4,
"suicide_threshold":50,
"mem_for_compute":null,
"mem_for_management":null,
"sherlock_rest_port":29446,
"nfs_mount_options":"nolock,rw,timeo=60",
"rediscover_timeout":120,
"kvm_emulatorpin_cpuset":"",
"rdvm_bmapcache_invalidate_all":0,
"consul_serflan_port":29460,
"monitor_host_var_robin_volume_lowmark":0.85,
"rest_port":29450,
"monitor_report_interval":5,
"host_type":"physical",
"monitor_host_swap_highmark":0.8,
"nodejs_port":29447,
"monitor_push_interval":60,
"ovs_enabled":true,
"monitor_host_var_log_volume_highmark":0.9,
"kubelet_restart_tolerance":15,
"monitor_host_mem_highmark":0.85,
"monitor_host_var_log_volume_lowmark":0.85,
"log_level":10,
"monitor_host_var_crash_volume_lowmark":0.85,
"monitor_container_volume_highmark":0.9,
"monitor_container_volume_lowmark":0.85,
"consul_server_port":29459,
"monitor_host_var_pgsql_volume_highmark":0.7
},
"ssd_lused":0,
"hugepages_1g_allocated":0,
"k8s_node_status":"Ready",
"hdd_free_alloc_slices":293131517952.0,
"ssd_lalloc":0,
"hostname":"cscale-82-140.robinsystems.com",
"public_hostname":"cscale-82-140.robinsystems.com",
"hdd_max_alloc_slices":328564998144.0,
"memory_total":33555709952,
"mem_for_compute":4294967296,
"sysinfo":{
"join_time":1596576678,
"current_version":"5.3.0-171",
"iqn":"iqn.1994-05.com.redhat:329b8568de1",
"install_date":"Tue Mar 17 23:49:17 UTC 2020",
"wwpns":[
],
"distribution":"CentOS Linux",
"version":"#1 SMP Tue Mar 17 23:49:17 UTC 2020",
"uuid":"",
"boot_time":1596576222,
"robin_software":[
{
"version":"5.3.0",
"patch":"",
"full_version":"5.3.0-171",
"install_date":"2020-08-03",
"patch_date":"",
"release":"171",
"build_info":"robin-c2edf85eaa83a42ced9512e7de9c7c2f1e4fa962:robin-ui:9ee33fd00273ba19861d4dc3ef8c6169d822d3e0:robingraph:cf0ceefe696ccac2dbd2eeb1d28b859955452843"
}
],
"release":"3.10.0-1062.18.1.el7.x86_64",
"system":"Linux",
"processor":"x86_64"
},
"hdd_pused":1174405120,
"disks":[
{
"spf":0.8,
"zoneid":1596601846,
"dev":"\/dev\/sda",
"aslices":0,
"nodeid":1,
"maintenance_mode":"OFF",
"role":"RootDisk",
"protected":0,
"status":"UNKNOWN",
"make":null,
"reattachable_nodes":[
[
"cscale-82-140.robinsystems.com",
"ONLINE"
]
],
"capacity":107374182400,
"max_latency_sensitive_vols_per_disk":2,
"pfree":0,
"node_hostname":"cscale-82-140.robinsystems.com",
"tags":{
},
"pused":0,
"type":"HDD",
"nvols":0,
"state":"INIT",
"reattachpolicy":{
"restarts_done":0,
"burst_count":0,
"burst_start_time":0,
"burst_interval":600,
"id":1,
"restart_limit":5
},
"max_alloc_slices":77,
"stormgrid":0,
"free_alloc_slices":77,
"slices":0,
"availability_zone":null,
"max_throughput_intensive_vols_per_disk":1,
"model":null,
"lused_size":0,
"devpath":"\/dev\/disk\/by-id\/scsi-3600224801d3ac9b6650afd3280aa5898",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"wwn":"0x600224801d3ac9b6650afd3280aa5898",
"allocations":[
],
"alloc_score":0,
"node_ref":1,
"preserved":0
},
{
"spf":0.8,
"zoneid":1596601846,
"dev":"\/dev\/dm-0",
"aslices":0,
"nodeid":1,
"maintenance_mode":"OFF",
"role":"RootDisk",
"protected":0,
"status":"UNKNOWN",
"make":null,
"reattachable_nodes":[
[
"cscale-82-140.robinsystems.com",
"ONLINE"
]
],
"capacity":53687091200,
"max_latency_sensitive_vols_per_disk":2,
"pfree":0,
"node_hostname":"cscale-82-140.robinsystems.com",
"tags":{
},
"pused":0,
"type":"HDD",
"nvols":0,
"state":"INIT",
"reattachpolicy":{
"restarts_done":0,
"burst_count":0,
"burst_start_time":0,
"burst_interval":600,
"id":4,
"restart_limit":5
},
"max_alloc_slices":38,
"stormgrid":0,
"free_alloc_slices":38,
"slices":0,
"availability_zone":null,
"max_throughput_intensive_vols_per_disk":1,
"model":null,
"lused_size":0,
"devpath":"\/dev\/disk\/by-id\/dm-uuid-LVM-vI83PDTxV3H0dWyAXfH5ef7rxTOuYyphgpZcvqGdfOKaXbEbOZzNthc6btsoSXDj",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"wwn":"0x600224801d3ac9b6650afd3280aa5898-centos-root",
"allocations":[
],
"alloc_score":0,
"node_ref":1,
"preserved":0
},
{
"spf":0.8,
"zoneid":1596601846,
"dev":"\/dev\/dm-1",
"aslices":0,
"nodeid":1,
"maintenance_mode":"OFF",
"role":"RootDisk",
"protected":0,
"status":"UNKNOWN",
"make":null,
"reattachable_nodes":[
[
"cscale-82-140.robinsystems.com",
"ONLINE"
]
],
"capacity":8254390272,
"max_latency_sensitive_vols_per_disk":2,
"pfree":0,
"node_hostname":"cscale-82-140.robinsystems.com",
"tags":{
},
"pused":0,
"type":"HDD",
"nvols":0,
"state":"INIT",
"reattachpolicy":{
"restarts_done":0,
"burst_count":0,
"burst_start_time":0,
"burst_interval":600,
"id":5,
"restart_limit":5
},
"max_alloc_slices":5,
"stormgrid":0,
"free_alloc_slices":5,
"slices":0,
"availability_zone":null,
"max_throughput_intensive_vols_per_disk":1,
"model":null,
"lused_size":0,
"devpath":"\/dev\/disk\/by-id\/dm-uuid-LVM-vI83PDTxV3H0dWyAXfH5ef7rxTOuYyphaFy4aq3EUo1yluonS8FG0LF16ycBrdEw",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"wwn":"0x600224801d3ac9b6650afd3280aa5898-centos-swap",
"allocations":[
],
"alloc_score":0,
"node_ref":1,
"preserved":0
},
{
"spf":0.8,
"zoneid":1596601846,
"dev":"\/dev\/dm-2",
"aslices":0,
"nodeid":1,
"maintenance_mode":"OFF",
"role":"RootDisk",
"protected":0,
"status":"UNKNOWN",
"make":null,
"reattachable_nodes":[
[
"cscale-82-140.robinsystems.com",
"ONLINE"
]
],
"capacity":44350570496,
"max_latency_sensitive_vols_per_disk":2,
"pfree":0,
"node_hostname":"cscale-82-140.robinsystems.com",
"tags":{
},
"pused":0,
"type":"HDD",
"nvols":0,
"state":"INIT",
"reattachpolicy":{
"restarts_done":0,
"burst_count":0,
"burst_start_time":0,
"burst_interval":600,
"id":6,
"restart_limit":5
},
"max_alloc_slices":32,
"stormgrid":0,
"free_alloc_slices":32,
"slices":0,
"availability_zone":null,
"max_throughput_intensive_vols_per_disk":1,
"model":null,
"lused_size":0,
"devpath":"\/dev\/disk\/by-id\/dm-uuid-LVM-vI83PDTxV3H0dWyAXfH5ef7rxTOuYyphQObDlS6eMUSpSxH5zsvyg9I5a0Gpuj5W",
"alloc_slices":0,
"reattachable":0,
"max_volumes_per_disk":10,
"wwn":"0x600224801d3ac9b6650afd3280aa5898-centos-home",
"allocations":[
],
"alloc_score":0,
"node_ref":1,
"preserved":0
},
{
"spf":0.8,
"zoneid":1596601846,
"dev":"\/dev\/sdb",
"aslices":7,
"nodeid":1,
"maintenance_mode":"OFF",
"role":"Storage",
"write_unit":4096,
"status":"ONLINE",
"make":null,
"reattachable_nodes":[
[
"cscale-82-140.robinsystems.com",
"ONLINE"
]
],
"protected":0,
"capacity":107374182400,
"max_latency_sensitive_vols_per_disk":2,
"pfree":104287174656,
"node_hostname":"cscale-82-140.robinsystems.com",
"tags":{
},
"pused":234881024,
"type":"HDD",
"nvols":3,
"state":"READY",
"reattachpolicy":{
"restarts_done":0,
"burst_count":0,
"burst_start_time":0,
"burst_interval":600,
"id":2,
"restart_limit":5
},
"max_alloc_slices":77,
"stormgrid":1,
"free_alloc_slices":68,
"slices":6390,
"availability_zone":null,
"max_throughput_intensive_vols_per_disk":1,
"model":null,
"lused_size":0,
"devpath":"\/dev\/disk\/by-id\/scsi-3600224804c48fd7e16c608dea0919064",
"alloc_slices":9,
"reattachable":0,
"max_volumes_per_disk":10,
"wwn":"0x600224804c48fd7e16c608dea0919064",
"allocations":[
{
"vols":[
{
"media":"HDD",
"pused":167772160,
"id":"1",
"size":5368709120,
"state":"ONLINE",
"name":"file-collection-1596578146092.269f9b38-f828-48c2-a382-8921dd74ee53"
}
],
"volume_group":"file-collection-1596578146092.269f9b38-f828-48c2-a382-8921dd74ee53.72.1.673abece-0975-4234-9fc2-56a06bf54031",
"name":"file-collection-1596578146092.269f9b38-f828-48c2-a382-8921dd74ee53.0.970a44c7-a15c-4612-ac57-9b4f15ae386e",
"volume":{
"media":"HDD",
"pused":167772160,
"id":"1",
"size":5368709120,
"state":"ONLINE",
"name":"file-collection-1596578146092.269f9b38-f828-48c2-a382-8921dd74ee53"
},
"slices":5
},
{
"vols":[
{
"media":"HDD",
"pused":67108864,
"id":"8",
"size":1073741824,
"state":"ONLINE",
"name":"test-RIC-1.server.01.data.1.382f1ad5-1294-4e24-8297-9c6025eacfe5"
}
],
"volume_group":"test-RIC-1.server.01.72.1.ea297971-f931-4787-99cc-6782e026b77c",
"name":"test-RIC-1.server.01.72.1.ea297971-f931-4787-99cc-6782e026b77c.0.1e8feb41-fd42-4409-b8c1-751331febdc1",
"volume":{
"media":"HDD",
"pused":67108864,
"id":"8",
"size":1073741824,
"state":"ONLINE",
"name":"test-RIC-1.server.01.data.1.382f1ad5-1294-4e24-8297-9c6025eacfe5"
},
"slices":2
},
{
"vols":[
{
"media":"HDD",
"pused":0,
"id":"9",
"size":1073741824,
"state":"ONLINE",
"name":"test-RIC-1.server.01.block.1.1053eaeb-4542-42a5-a173-d69a76703ead"
}
],
"volume_group":"test-RIC-1.server.01.72.1.1e483fd0-2d5c-434c-aef0-91a87796977a",
"name":"test-RIC-1.server.01.72.1.1e483fd0-2d5c-434c-aef0-91a87796977a.0.5bf728c4-ea26-4e0f-82d7-c584fcf0bd9a",
"volume":{
"media":"HDD",
"pused":0,
"id":"9",
"size":1073741824,
"state":"ONLINE",
"name":"test-RIC-1.server.01.block.1.1053eaeb-4542-42a5-a173-d69a76703ead"
},
"slices":2
}
],
"alloc_score":95,
"node_ref":1,
"preserved":0
},
{
"spf":0.8,
"zoneid":1596601846,
"dev":"\/dev\/sdc",
"aslices":20,
"nodeid":1,
"maintenance_mode":"OFF",
"role":"Storage",
"write_unit":4096,
"status":"ONLINE",
"make":null,
"reattachable_nodes":[
[
"cscale-82-140.robinsystems.com",
"ONLINE"
]
],
"protected":0,
"capacity":107374182400,
"max_latency_sensitive_vols_per_disk":2,
"pfree":103582531584,
"node_hostname":"cscale-82-140.robinsystems.com",
"tags":{
},
"pused":939524096,
"type":"HDD",
"nvols":1,
"state":"READY",
"reattachpolicy":{
"restarts_done":0,
"burst_count":0,
"burst_start_time":0,
"burst_interval":600,
"id":3,
"restart_limit":5
},
"max_alloc_slices":77,
"stormgrid":2,
"free_alloc_slices":53,
"slices":6390,
"availability_zone":null,
"max_throughput_intensive_vols_per_disk":1,
"model":null,
"lused_size":0,
"devpath":"\/dev\/disk\/by-id\/scsi-3600224803bcdafde95b1f5cd27ceb5fb",
"alloc_slices":24,
"reattachable":0,
"max_volumes_per_disk":10,
"wwn":"0x600224803bcdafde95b1f5cd27ceb5fb",
"allocations":[
{
"vols":[
{
"media":"HDD",
"pused":939524096,
"id":"16",
"size":21474836480,
"state":"ONLINE",
"name":"rohan-app.nginx.03.data.1.83d03fbf-3bfe-4723-8abe-5cbd51014e0c"
}
],
"volume_group":"rohan-app.nginx.03.72.1.44251a23-0221-4fda-837e-db26bca3ccb8",
"name":"rohan-app.nginx.03.72.1.44251a23-0221-4fda-837e-db26bca3ccb8.0.6e2afb65-5c4c-42e4-972e-161de3fb3856",
"volume":{
"media":"HDD",
"pused":939524096,
"id":"16",
"size":21474836480,
"state":"ONLINE",
"name":"rohan-app.nginx.03.data.1.83d03fbf-3bfe-4723-8abe-5cbd51014e0c"
},
"slices":24
}
],
"alloc_score":89,
"node_ref":1,
"preserved":0
}
]
}
]
}
4.6. Updating node attributes¶
Robin allows a user to update the CPU provisioning factor of a particular host for non-isolated cores as well as isolated cores that are shared. The default value for this attribute is 10. This implies each CPU on the node is overprovisioned by a factor of 10. The reason for this is that CPUs are generally the most underutilized resource in a datacenter. However for environments wherein which highly CPU intensive workloads are intended to run, the CPU provisioning factor can be set to 1 which entails strictly no overprovisioning should occur.
In order to update the CPU provisioning factor, issue the following command:
# robin host update [<hostname>]
--cpu-provisioning-factor <factor>
--all
|
FQDN of host |
|
CPU provisioning factor to update to |
|
Run update for all hosts. Should be specified if no hostname is given |
Example:
# robin host update vnode36.robinsystems.com --cpu-provisioning-factor 1 --wait
Job: 226 Name: HostModify State: PROCESSED Error: 0
Job: 226 Name: HostModify State: COMPLETED Error: 0
Updates the CPU provisioning factor for a particular host in order to help tune it for the different types of workloads intended to run on it.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: update- This mandatory field within the payload specifies the update action is to be performed.cpuprovfactor: <prov_factor>- This mandatory field within the payload specifies the CPU provisioning factor to be set for the aforementioned host.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":89
}
4.7. Tagging a node¶
Whilst Robin provides host isolation/grouping at a higher level via a construct called resource pools (more information detailed here), a more granular control
is sometimes needed. In order to achieve this Robin allows users to tag hosts with user defined tags which not only allows for separation within a resource pool but also enables
the user to take advantage of Robin’s advanced application policies and enforce exactly which host(s) an application can be deployed on. This granular control is useful as it allows for on
the fly alignment of resources and provide a temporary boundary for resources in comparison to resource pools which are more static in nature.
Tags are defined by key-value pairs and need to be registered with Robin before they can be assigned to hosts.
The following commands are described in this section:
|
Register a tag |
|
Assign a tag to a host |
|
Unassign a tag from a host |
|
Unregister a tag |
4.7.1. Registering a tag¶
In order to add a tag with Robin such that it can be assigned to a host, issue the following command:
# robin tag add-key [<key>] [<values>]
Note
If a key already exists, the above command can be used to append values for the aforementioned key
|
Key to be used for tag |
|
Space separated list of values to be associated with key |
Example:
# robin tag add-key color blue red
Added key 'color' with values: ['red', 'blue']
Registers a tag with Robin such that it can be assigned to a host.
End Point: /api/v3/robin_server/tags/
Method: PUT
URL Parameters: None
Data Parameters:
action: addkey- This mandatory field within the payload specifies the add key action is to be performed.key: <key_value>- This mandatory field within the payload specifies the key to be used for the tag.values: <list_of_values>- This mandatory field within the payload specifies a list of comma seperated strings to be used as the values associated with the aforementioned key.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 401 (Unauthorized Error)
Example Response:
Output
{
"message":"Added the following values: red, blue for key 'color'"
}
4.7.2. Assigning a tag to a node¶
In order to assign one or more tags to a host, issue the following command:
# robin host add-tags [<hostname>] [<tags>]
Note
Only one value for one key can be assigned to each host at any given moment.
|
FQDN of host |
|
Tags to assign in <key>=<value format |
Example:
# robin host add-tags vnode36.robinsystems.com color=blue
Job: 80 Name: SetTag State: VALIDATED Error: 0
Job: 80 Name: SetTag State: COMPLETED Error: 0
Assigns one or more tags to a host.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: add_tag- This mandatory field within the payload specifies the add tag action is to be performed.tags: <dict_of_key_value_pairs>–<tag_key>: <list_of_tag_values>- Each key value pair should be the tag’s key and a list of comma seperated values already associated with it. An example would be"color":["blue"]
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 401 (Unauthorized Error), 404 (Not Found Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid": 27
}
4.7.3. Unassigning a tag from a node¶
In order to unassign one or more tags to a host, issue the following command:
# robin host del-tags [<hostname>] [<tags>]
|
FQDN of host |
|
Tags to unassign in <key>=<value> format |
Example:
# robin host del-tags vnode36.robinsystems.com color=blue
Job: 81 Name: RemoveTag State: PROCESSED Error: 0
Job: 81 Name: RemoveTag State: COMPLETED Error: 0
Unassigns one or more tags from a host.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: del_tag- This mandatory field within the payload specifies the delete tag action is to be performed.tags: <dict_of_key_value_pairs>-<tag_key>: <list_of_tag_values>- Each key value pair should be the tag’s key and a list of comma seperated values already associated with it. An example would be"color":["blue"]
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 401 (Unauthorized Error), 404 (Not Found Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid": 28
}
4.7.4. Unegistering a tag¶
In order to remove a tag from Robin, issue the following command:
# robin tag del-key [<key>] [<values>]
Note
In order to delete a key, all of its associated values must be specified.
|
Key of tag to be deleted |
|
Space separated list of values to be unassociated with the key |
Example:
# robin tag del-key color blue red
Deleted tags with key 'color' for the following values: 'red, blue'
Removes a tag registered with Robin or disassociates certain values from a key.
End Point: /api/v3/robin_server/tags/
Method: PUT
URL Parameters: None
Data Parameters:
action: delkey- This mandatory field within the payload specifies the delete key action is to be performed.key: <key_value>- This mandatory field within the payload specifies the key of the tag to be removed/edited.values: <list_of_values>- This mandatory field within the payload specifies a list of comma seperated strings to be disassociated with the aforementioned key. If all values associated with the key are specified, the key is removed and the tag is deleted.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 401 (Unauthorized Error), 404 (Not Found Error)
Example Response:
Output
{
"message":"Deleted tags with key 'color' for the following values: 'blue'"
}
4.8. Disabling a node¶
In certain situations, a user might not want any resources (storage or compute) for an application to be allocated from a particular host due to a malfunction with the physical machine or simply because the host is temporarily undergoing maintenance. Instead of requiring the user to remove the node from an existing cluster, Robin allows one to place a host into maintenance mode. This effectively isolates the host with regards to resource availability as it entails that none of the host’s storage capacity or compute resources can be used for future application deployment regardless of the Robin roles assigned to the node. This mode can be toggled using the commands detailed below. For more granular control, please review the section on disabling/enabling particular roles here.
The following commands are described in this section:
|
Place a host into maintenance mode |
|
Place a host into non-maintenance (normal) mode |
4.8.1. Placing a host into maintenance mode¶
In order to put a host into maintenance and thus temporarily suspend it from providing either storage or compute resources for future application deployments, issue the following command:
# robin host set-maintenance <hostname>
|
FQDN of host |
Example:
# robin host set-maintenance vnode36.robinsystems.com
Host vnode36.robinsystems.com set in maintenance mode
Puts a host into maintenance mode, which in turn temporarily suspends it from providing storage and compute resources for application deployments.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: set_maintenance- This mandatory field within the payload specifies that the set maintenance mode operation is to be performed.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"message":"Maintenance mode set"
}
4.8.2. Placing a host into non-maintenance mode¶
In order to revert a host back into its normal setting and thus allow it to provide resources for future application deployments, issue the following command:
# robin host unset-maintenance <hostname>
|
FQDN of host |
Example:
# robin host unset-maintenance vnode36.robinsystems.com
Host vnode36.robinsystems.com out of maintenance mode
Removes a host from maintenance mode, which in turn allows it to provide storage and compute resources for application deployments.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: unset_maintenance- This mandatory field within the payload specifies that the unset maintenance mode operation is to be performed.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 200
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"message":"Maintenance mode unset"
}
4.9. Decommissioning a node¶
When a server physically malfunctions or needs to be replaced, one has to remove the node from its respective Robin and Kubernetes clusters. Detailed below is a step-by-step walkthrough which showcases how to achieve this, whilst also ensuring your Robin cluster continues to function normally.
Removing roles from a host:
Before a host can be removed from the Robin cluster all the roles currently assigned to it must be removed. Detailed below are the different preconditions that need to be met before removing a particular role.
If the host in question has the Compute role assigned to it, all pods from this host will have to be relocated to other hosts in the cluster. This can be achieved by running the robin instance relocate command
for all the Vnodes that are currently residing on the host. More details about this command can be found here. The reason for this is that the Compute role denotes that resources such as CPU and memory can be allocated from this host. As a result in order to remove
the role no compute resources should be taken from the host. This is the resulting consequence of moving the respective pods over.
# robin instance list --host intel-1.robinsystems.com
ID | Owner/Tenant | ResPool | Container | Hostname | IP Addresses | Status | Autopilot | LastOpr | Host | Engine | Cores | GPUs | Mem (GB) | Ctime
---+----------------------+---------+--------------+---------------------------------------------+----------------+--------+-----------+---------+---------+--------+-------+------+----------+----------------------
1 | robin/Administrators | default | a1.server.01 | a1-server-01.t001-u000003.svc.cluster.local | 172.21.17.3/16 | ONLINE | Enabled | STARTED | intel-1 | docker | 1 | 0 | 0.38 | 27 Feb 2020 13:40:13
# robin instance relocate a1.server.01 --wait --yes
Job: 63 Name: VnodeDeploy State: PROCESSED Error: 0
Job: 63 Name: VnodeDeploy State: WAITING Error: 0
Job: 63 Name: VnodeDeploy State: AGENT_WAIT Error: 0
Job: 63 Name: VnodeDeploy State: FINALIZED Error: 0
Job: 63 Name: VnodeDeploy State: COMPLETED Error: 0
# robin instance list
ID | Owner/Tenant | ResPool | Container | Hostname | IP Addresses | Status | Autopilot | LastOpr | Host | Engine | Cores | GPUs | Mem (GB) | Ctime
---+----------------------+---------+--------------+---------------------------------------------+----------------+--------+-----------+---------+--------------+--------+-------+------+----------+----------------------
1 | robin/Administrators | default | a1.server.01 | a1-server-01.t001-u000003.svc.cluster.local | 172.21.17.3/16 | ONLINE | Enabled | STARTED | cscale-82-83 | docker | 1 | 0 | 0.38 | 27 Feb 2020 13:40:13
If the host in question has the Storage role assigned to it, all
volumes that are currently allocated on drives which belong to that
host must be evacuated to drives residing on other hosts. This can be
achieved by running the robin drive evacuate command. More
details about this command can be found here. The reason
for this is that the Storage role denotes that the host is available
to provide storage capacity for volumes that are created alongisde
applications. As a result, in order to remove the role there must be
no volume allocations tied with the host. This is the resulting
consequence of evacuating all the volumes from each drive on the
node.
# robin drive list --host intel-1.robinsystems.com
ID | WWN | Host | Path /dev/disk/by-id | Size(GB) | Movable | Type | Free/Max(GB) | Vols | Role | Status | LastOpr
---+-------------------------------------------+---------+-----------------------------------------------+----------+---------+------+--------------+------+---------+--------+---------
3 | 0xQEMU_QEMU_HARDDISK_e54d6149-0a4e-48ce-b | intel-1 | scsi-0QEMU_QEMU_HARDDISK_e54d6149-0a4e-48ce-b | 100 | N | HDD | 63/77 (82%) | 2/10 | Storage | ONLINE | READY
4 | 0xQEMU_QEMU_HARDDISK_89fc0488-2050-4f44-a | intel-1 | scsi-0QEMU_QEMU_HARDDISK_89fc0488-2050-4f44-a | 100 | N | HDD | 77/77 (100%) | 0/10 | Storage | ONLINE | READY
Only storage disks are shown. Issue `robin disk list --role all` to view all disks
# robin drive evacuate 0xQEMU_QEMU_HARDDISK_e54d6149-0a4e-48ce-b --exclude-disks 0xQEMU_QEMU_HARDDISK_e54d6149-0a4e-48ce-b, 0xQEMU_QEMU_HARDDISK_89fc0488-2050-4f44-a --wait --yes
Job: 65 Name: DiskEvacuate State: VALIDATED Error: 0
Job: 65 Name: DiskEvacuate State: WAITING Error: 0
Job: 65 Name: DiskEvacuate State: COMPLETED Error: 0
# robin drive list --host intel-1.robinsystems.com
ID | WWN | Host | Path /dev/disk/by-id | Size(GB) | Movable | Type | Free/Max(GB) | Vols | Role | Status | LastOpr
---+-------------------------------------------+---------+-----------------------------------------------+----------+---------+------+--------------+------+---------+--------+---------
3 | 0xQEMU_QEMU_HARDDISK_e54d6149-0a4e-48ce-b | intel-1 | scsi-0QEMU_QEMU_HARDDISK_e54d6149-0a4e-48ce-b | 100 | N | HDD | 77/77 (100%) | 0/10 | Storage | ONLINE | READY
4 | 0xQEMU_QEMU_HARDDISK_89fc0488-2050-4f44-a | intel-1 | scsi-0QEMU_QEMU_HARDDISK_89fc0488-2050-4f44-a | 100 | N | HDD | 77/77 (100%) | 0/10 | Storage | ONLINE | READY
If the host in question has the Manager role assigned to it, there are no explicit preconditions that need to be met. However we recommend that for high availability clusters that there are at least 2 nodes with the Manager role assigned to them at all times. This is because if the master Manager node does fail or is rebooted, a cluster failover can occur and the Robin cluster is not affected.
Once all the necessary conditions have been met, the roles can be removed as shown in the following example:
# robin host list
Id | Hostname | Version | Status | LastOpr | Resource Pool | Roles | Cores | GPUs | Mem(Free/Alloc/Total) | HDD(#/Alloc/Total) | SSD(#/Alloc/Total) | Instances | Joined Time
-------------+-------------------------------+------------+--------+---------+---------------+-------+-------+------+-----------------------+--------------------+--------------------+-----------+----------------------
1582820722:1 | cscale-82-81.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M* | 2/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:25:46
1582820722:2 | cscale-82-82.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:32:47
1582820722:3 | cscale-82-83.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:41:56
1582820722:4 | intel-1.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | S,C | 1/720 | 0/0 | 175G/12G/187G | 2/-/200G | -/-/- | 0 | 27 Feb 2020 00:53:35
# robin host remove-role intel-1.robinsystems.com storage,compute --yes --wait
Job: 208 Name: HostRemoveRoles State: VALIDATED Error: 0
Job: 208 Name: HostRemoveRoles State: WAITING Error: 0
Job: 208 Name: HostRemoveRoles State: COMPLETED Error: 0
# robin host list
Id | Hostname | Version | Status | LastOpr | Resource Pool | Roles | Cores | GPUs | Mem(Free/Alloc/Total) | HDD(#/Alloc/Total) | SSD(#/Alloc/Total) | Instances | Joined Time
-------------+-------------------------------+------------+--------+---------+---------------+-------+-------+------+-----------------------+--------------------+--------------------+-----------+----------------------
1582820722:1 | cscale-82-81.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M* | 2/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:25:46
1582820722:2 | cscale-82-82.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:32:47
1582820722:3 | cscale-82-83.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M,C | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:41:56
1582820722:4 | intel-1.robinsystems.com | 5.2.1-9769 | Ready | SYNCED | default | | 1/720 | 0/0 | 187G/0.05G/187G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:53:35
Removing the host:
Once all the the roles are removed from a host, issue the following command to remove the host from a Robin cluster:
# robin host remove [<host>]
--force
--yes
|
FQDN of host |
|
Forcibly remove a host from a cluster. Required if removing a node that is down |
|
Do not prompt the user for confirmation of removal |
Example:
# robin host list
Id | Hostname | Version | Status | LastOpr | Resource Pool | Roles | Cores | GPUs | Mem(Free/Alloc/Total) | HDD(#/Alloc/Total) | SSD(#/Alloc/Total) | Instances | Joined Time
-------------+-------------------------------+------------+--------+---------+---------------+-------+-------+------+-----------------------+--------------------+--------------------+-----------+----------------------
1582820722:1 | cscale-82-81.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M* | 2/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:25:46
1582820722:2 | cscale-82-82.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:32:47
1582820722:3 | cscale-82-83.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:41:56
1582820722:4 | intel-1.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | S,C | 1/720 | 0/0 | 175G/12G/187G | 2/-/200G | -/-/- | 0 | 27 Feb 2020 00:53:35
# kubectl get nodes
NAME STATUS ROLES AGE VERSION
cscale-82-81 Ready master 21h v1.16.3
cscale-82-82 Ready master 21h v1.16.3
cscale-82-83 Ready master 21h v1.16.3
intel-1 Ready worker 21h v1.16.3
# robin host remove intel-1.robinsystems.com --yes --wait
Job: 215 Name: HostRemove State: VALIDATED Error: 0
Job: 215 Name: HostRemove State: WAITING Error: 0
Job: 215 Name: HostRemove State: COMPLETED Error: 0
# robin host list
Id | Hostname | Version | Status | LastOpr | Resource Pool | Roles | Cores | GPUs | Mem(Free/Alloc/Total) | HDD(#/Alloc/Total) | SSD(#/Alloc/Total) | Instances | Joined Time
-------------+-------------------------------+------------+--------+---------+---------------+-------+-------+------+-----------------------+--------------------+--------------------+-----------+----------------------
1582820722:1 | cscale-82-81.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M* | 2/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:25:46
1582820722:2 | cscale-82-82.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:32:47
1582820722:3 | cscale-82-83.robinsystems.com | 5.2.1-9769 | Ready | ONLINE | default | M | 1/400 | 0/0 | 30G/1G/31G | -/-/- | -/-/- | 0 | 27 Feb 2020 00:41:56
# kubectl get nodes
NAME STATUS ROLES AGE VERSION
cscale-82-81 Ready master 21h v1.16.3
cscale-82-82 Ready master 21h v1.16.3
cscale-82-83 Ready master 21h v1.16.3
4.10. Managing a cluster via the remote client¶
In addition to the Robin CLI, which is available on all hosts where Robin is installed, a remote client is shipped with each cluster that is deployed.
This client mirrors the functionality of the native CLI with regards to the commands available and hence it provides the management capabilities that are described
throughout this document. One advantage of utilizing this client is that it can be used to manage a multitude of Robin clusters via the concept of contexts.
A context in this scenario refers to a Robin cluster and is identified by the server name or IP Address. In addition to this primary key, the following attributes
can also be set within a context: the port values for various Robin services (including the Robin Server, File Server, Event Server, Watchdog Server, and Metrics Server)
along with the logging level. The attributes are discussed in more detail in the following sections. After creating the appropriate context for a Robin cluster,
one can set it to be the current context and communicate with the respective cluster. The commands which can be used to achieve this are described below.
The following commands are described in this section:
|
Add a Robin cluster context |
|
List all registered Robin cluster contexts |
|
Set a Robin cluster context as the current context |
|
Update attributes for the current Robin cluster context |
|
Delete a Robin cluster context |
4.10.1. Downloading the Robin client¶
In order to download the Robin client from an existing Robin cluster, issue the following command:
# curl -k https://<master_ip>:<port>/api/v3/robin_server/download?file=robincli&os=<os> -o robin
|
IP Address of the Master Node or VIP |
|
Port number for the Robin Server |
|
The operating system to download the client for. Supported operating systems include: Linux, MacOS. |
Example:
# curl -k 'https://vnode42:29442/api/v3/robin_server/download?file=robincli&os=linux' -o robin
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 10.1M 100 10.1M 0 0 1421k 0 0:00:07 0:00:07 --:--:-- 1483k
# ls -lart
-rw-r--r-- 1 demo staff 10655536 Mar 26 14:12 robin
4.10.2. Adding a Context¶
A context is a construct that can be used to define a Robin cluster in a manner that the remote client can understand. In order to add a context, issue the following command:
Note
If a context already exists with the server specified, that context will be updated with the values supplied.
# robin client add-context <server>
--port <port>
--file-port <file_port>
--event-port <event_port>
--watchdog-port <watch_port>
--metrics-port <metrics_port>
--log-level <log_level>
--product <product_type>
--set-current
|
FQDN/IP Address of the Master Node or VIP |
|
Port number for the Robin Server. Default value is 29442 |
|
Port number for the File Server. Default value is 29445 |
|
Port number for the Event Server. Default value is 29449 |
|
Port number for the Watchdog Server. Default value is 29444 |
|
Port number for the Metrics Server. Default is 29446 |
|
Number indicating the verbosity of logs. Valid values are 10 (DEBUG), 20 (INFO), 40 (ERROR). Default value is 40. |
|
Type of ROBIN installation. Valid choices are ‘platform’ or ‘storage’. Default value is ‘platform’. |
|
Set context to be created as the current |
Note
If the target Robin cluster was deployed as a highly available cluster, ensure that the robincp_mode config attribute is enabled and set the port value for the Robin, file and event server to 29465. In addition the server parameter should be set to the value of the VIP in this case. Otherwise the remote client will not be fully operational.
Example:
# robin client add-context centos-60-214 --port 29442
Context robin-cluster-centos-60-214 created successfully
4.10.3. Listing all available contexts¶
In order to list all contexts that have already been registered with the client alongside additional details such as the port values specified or the log level, issue the following command:
# robin client list-contexts --full
|
Show additional details about all registered contexts |
Example:
# robin client list-contexts --full
| Server | Port | Version | Tenant | Last Login | Tenants | FPort | WPort | MPort | LogLevel
---+-----------------------------------+-------+------------+----------------+----------------------+----------------+-------+-------+-------+----------
| master.robin-server.service.robin | 29442 | - | - | - | | 29445 | 29444 | 29446 | ERR
| centos-60-214 | 29443 | - | Administrators | - | | 29445 | 29444 | 29446 | ERR
* | 172.19.174.194 | 29442 | 5.2.3-9842 | Administrators | 26 Mar 2020 16:10:58 | Administrators | 29445 | 29444 | 29446 | ERR
Note
The asterisk displayed above indicates the current context.
4.10.4. Setting the current context¶
In order to access a particular Robin cluster, its respective context needs to be set as the current context. To achieve this, issue the following command:
# robin client set-current <context>
|
The server attribute of the context to be set as current |
Example:
# robin client set-current centos-60-214
Current context set to robin-cluster-centos-60-214
4.10.5. Updating the current context¶
In certain situations, such as a reinstallation, the attributes of a context might be altered whilst retaining the same server IP Address or hostname. As a result, the context which refers to this cluster will have to be updated. In order to do so, issue the following command:
Note
The below command only updates the current context.
# robin client update-context --port <port>
--file-port <file_port>
--event-port <event_port>
--watchdog-port <watch_port>
--metrics-port <metrics_port>
--log-level <log_level>
|
Updated port number for the Robin Server |
|
Updated port number for the File Server |
|
Updated port number for the Event Server |
|
Updated port number for the Watchdog Server |
|
Updated port number for the Metrics Server |
|
Updated number indicating the verbosity of logs. Valid values are 10 (DEBUG), 20 (INFO), 40 (ERROR) |
Example:
# robin client update-context --port 29942 --file-port 29445 --watchdog-port 29444 --metrics-port 29446
Updating attributes for context robin-cluster-centos-60-214
Server: centos-60-214
Context config updated for robin-cluster-centos-60-214
4.10.6. Deleting a context¶
In order to remove a registered context, issue the following command:
# robin client delete-context <context>
|
The server attribute of the context to be deleted |
Example:
# robin client delete-context centos-60-214
Context centos-60-214 deleted
4.11. Managing Unused Images¶
Unused docker images tend to bloat and consume system storage. Whilst they can be deleted, manually locating and removing them becomes cumbersome. Robin facilitates the deletion of unused images (which are not in use by any containers) on all hosts (or a given subset) within the cluster. In addition, users can schedule a job to prune unnecessary images so as to mimic garbage collection for these images.
The following commands are described in this section:
|
Prune unused images from hosts within the cluster |
4.11.1. Pruning Images¶
In order to remove unused images on a host, issue the following command:
# robin host pruneimage [<hostnames>]
--skip-images-file <skip_images_path>
--all
|
Comma seperated list of host FQDNs |
|
Path to file containing a list of images that should not be pruned. Each image should be specified on a seperate line and in the following format: <image>:<tag> |
|
Prune unused images from all hosts. Should be specified if no hostname is given |
Example:
# robin host pruneimage vnode36.robinsystems.com --wait
Job: 226 Name: HostPruneImagesMulti State: PROCESSED Error: 0
Job: 226 Name: HostPruneImagesMulti State: COMPLETED Error: 0
Removes unused images from a specified host.
End Point: /api/v3/robin_server/hosts/<hostname>
Method: PUT
URL Parameters: None
Data Parameters:
action: pruneimage- This mandatory field within the payload specifies that the image prune operation should be performed.skip_images: <list_of_imgs>- Utilizing this parameter within the payload, by specifying a list of images in the format<image_name>:<tag>, results in the given images not being removed even if they are unused at the moment.
Port: RCM Port (default value is 29442)
Headers:
Authorization: <auth_token>: Authorization token to identify which user is sending the request. The token can be acquired from the login API.
Success Response Code: 202
Error Response Code: 500 (Internal Server Error), 404 (Not Found Error), 401 (Unauthorized Error), 400 (Invalid API Usage Error)
Example Response:
Output
{
"jobid":1039
}
4.11.2. Enabling Pruning Schedule¶
In order to enable a schedule that removes all unused images via a CRON string, issue the following command:
# robin config update server image_prune_cron <cron_string>
|
CRON string representing the time at which the schedule shouild run |
Note
In order to set the schedule via a CRON string, it has to be enabled first via the following command: robin config update server enable_image_prune_schedule true
Example:
# robin config update server enable_image_prune_schedule true
The 'server' attribute 'enable_image_prune_schedule' has been updated
# robin config update server image_prune_cron "0 2 * * *"
The 'server' attribute 'image_prune_cron' has been updated
4.11.3. Disabling Pruning Schedule¶
In order to disable the schedule that removes all unused images, issue the following command:
# robin config update server enable_image_prune_schedule false
Example:
# robin config update server enable_image_prune_schedule false
The 'server' attribute 'enable_image_prune_schedule' has been updated