Cluster recovery manager to remediate failovers
Example implementations relate to management of clusters. A cluster recovery manager may comprise a processing resource; and a memory resource storing machine-readable instructions to cause the processing resource to: adjust, based on a monitored degree of performance of a controller of a controller cluster, a state of the controller to one of a first state and a second state; and reassign a corresponding portion of a plurality of APs managed by the controller periodically to a different controller until the state of the controller is determined to be adjustable to the first state. The reassignment can be triggered responsive to a state adjustment of the controller from the first state to the second state.
1. A cluster recovery manager, comprising:
a processing resource; and
a memory resource storing machine-readable instructions to cause the processing resource to:
determine a failed controller of a controller cluster;
responsive to the determination of the faded controller:
initiate an automatic recovery process of the failed controller that adjusts a slate of the failed controller from a first state to a second state, and
reassign a first portion of a plurality of access points (APs) managed by the failed controller to a second controller of the controller cluster at a first iteration of a periodic reassignment of the plurality of APs;
during a second iteration of the periodic reassignment of the plurality of APs that is different than the first iteration, reassign a second portion of the plurality of APs managed by the failed controller to a third controller of the controller cluster, wherein the second portion of the plurality of APs comprise different APs than the first portion; and
stop the periodic reassignment of the plurality of APs based on the automatic recovery process of the failed controller.
2. The cluster recovery manager of claim 1 , wherein the second portion of the plurality of APs is periodically reassignable in a respective period until the state of the failed controller is determined to be adjustable from the second state to the first state.
3. The cluster recovery manager of claim 1 , wherein the second state indicates a degree of performance of the failed controller below a threshold degree of performance.
4. The cluster recovery manager of claim 1 , wherein stopping the periodic assignment of the plurality of APs is also based on a determination that the state of the failed controller is adjustable from the second state to the first state.
5. The cluster recovery manager of claim 1 , wherein the automatic recovery process comprises periodic reassignments of user devices managed by the failed controller.
6. The cluster recovery manager of claim 1 , wherein the first state is normal and the second state is failure.
7. The cluster recovery manager of claim 1 , wherein the second controller is an access point anchor controller (active AAC) and the third controller is a user anchor controller (UAC).
8. A non-transitory machine-readable storage medium storing instructions executable by a processing resource to:
determine a failed controller of a controller cluster;
responsive to the determination of the faded controller:
initiate an automatic recovery process of the faded controller;
reassign a first portion of a plurality of access points (APs) managed by the failed controller periodically to first standby controllers of the controller cluster at a first iteration of a periodic reassignment; and
during a second iteration of the periodic reassignment of the plurality of APs that is different than the first iteration, reassign a second portion of the plurality of APs managed by the failed controller to second standby controllers of the controller cluster, wherein the second portion of the plurality of APs comprise different APs than the first portion; and
stop the periodic reassignment of the plurality of APs based on the automatic recovery process of the failed controller.
9. The medium of claim 8 , wherein determining the failed controller is based on information collected from the failed controller of the controller cluster, and wherein the information includes corresponding values of a plurality of parameters, and the instructions further to:
compare the plurality of parameters with a plurality of threshold values; and
determine, based on the comparison, whether a particular one of the plurality of parameters is below one of the plurality of threshold values.
10. The medium of claim 9 , wherein the information is collected periodically from the failed controller.
11. The medium of claim 8 , the instructions further to:
collect information associated with all controllers of the controller cluster; and
monitor respective degree of performances of all controllers of the controller cluster.
12. The medium of claim 8 , the instructions further to:
reassign a plurality of user devices managed by the failed controller, wherein the reassignment of the plurality of user devices is initiated when each of the plurality of APs is reassigned.
13. The medium of claim 8 , wherein each one of the first standby controllers is a standby access point anchor controller (S-AAC).
14. The medium of claim 8 , wherein each one of the second standby controllers is a standby user anchor controller (S-UAC).
15. The medium of claim 8 , wherein the automatic recovery process of the failed controller is to restart the failed controller.
16. A method, comprising:
determining a faded controller of a controller cluster;
responsive to the determination of the faded controller:
initiating an automatic recovery process of the failed controller;
reassigning, by a processing resource, a first portion of a plurality of access points (APs) managed by the failed controller of the controller cluster to a second controller of the controller cluster at a first iteration of a periodic reassignment of the plurality of APs;
during a second iteration of the periodic reassignment of the plurality of APs that is different than the first iteration, reassigning a second portion of the plurality of APs managed by the failed controller to a third controller of the controller cluster, wherein the second portion of the plurality of APs comprise different APs than the first portion; and
stopping the periodic reassignment of the plurality of APs based on the automatic recovery process of the failed controller.
17. The method of claim 16 , further comprising:
excluding, by the processing resource, the failed controller from the controller cluster.
18. The method of claim 17 , further comprising:
alerting, by the processing resource, an administrator that the failed controller is excluded from the controller cluster.
19. The method of claim 16 , wherein the first portion and the second portion of the plurality of APs are periodically reassigned prior to a plurality portions of user devices being periodically reassigned.
20. The method of claim 16 , further comprising:
waiting, subsequent to reassigning the first portion of the plurality of APs, a certain amount of time prior to reassigning the second portion of the plurality of APs.