Non-Disruptive Controller Scaling Troubleshooting

Everpure Cloud Dedicated for Azure

Audience
Public
Source Type
Documentation

This guides provides tips on how to troubleshoot the Non-Disruptive Controller Scaling procedure.

Context

While performing the Non-Disruptive Controller Scaling Procedure, one or more of the pre-upgrade checks may fail, resulting in the Controller upgrade to be paused with a warning.

To help with resolution, the system will identify the failed checks and offer possible next steps. While paused in this stage, the controller upgrade can be aborted. Alternatively, the failed health checks may be retried or overridden, allowing the upgrade to eventually proceed with the next steps.

Capacity Check Failed

In the below example, the capacity expansion has not been performed before starting the controller upgrade and the CapacityStepCheck failed. The upgrade process is paused, which is the default reaction to a failed check. In this case of insufficient capacity point, the only option is to abort the controller upgrade, perform capacity expansion, and start over.

purearray cloud-config model list
Current Model   	Requested Model 	Status  Step           	Details
CBS-V10MP2R2  CBS-V20MP2R2  paused  pre-upgrade-check  Upgrade has been paused: Finished pre-upgrade, 16 checks in total, 1 failed (CapacityStepCheck).
 
CapacityStepCheck
Message: Current disk capacity is not sufficient for upgrading to the CBS-V20MP2R2 model.
Recommended Action: Please abort the hardware upgrade, perform a capacity upgrade with a required step of at least 7696581394432, then start the hardware upgrade again.
Overridable: False
 
To proceed with the upgrade, either:
1) Retry upgrade checks - Run "purearray cloud-config model setattr --requested-model CBS-V20MP2R2 [--override <CheckName>]"
2) Abort the upgrade - Run "purearray cloud-config model setattr --requested-model ''"
3) Contact Pure Technical Services for assistance

Without aborting the controller upgrade, capacity expansion will not proceed:

purearray cloud-capacity setattr --requested-capacity 7696581394432
Error on capacity update: Cannot perform operation while hardware upgrade is in progress.

To abort the upgrade, run the following command:

purearray cloud-config model setattr --requested-model ''
Current Model  	Requested Model Status  Step  Details
CBS-V10MP2R2         	idle
Note: Other checks might allow you to retry or override them, without having to abort the entire process. However, the capacity check does not provide these options.

After you have completed the capacity expansion, you can re-initiate the upgrade process.

Overriding upgrade checks

Some of the checks can be overridden with the command --override <CheckName>.

The system will ignore the failing check and proceed with the upgrade. This is applicable not only to pre-upgrade checks but also for the subsequent mid-upgrade checks.

The below example shows two failed checks, the CapacityStepCheck which cannot be overridden (as indicated by the flag Overridable: False) and the HostPathsCheck which can be overridden (as indicated by the flag Overridable: True):

purearray cloud-config model list
Current Model   	Requested Model 	Status  Step           	Details
CBS-V10MP2R2  CBS-V20MP2R2  paused  pre-upgrade-check  Upgrade has been paused: Finished pre-upgrade, 16 checks in total, 2 failed (CapacityStepCheck, HostPathsCheck).
 
CapacityStepCheck
Message: Current disk capacity is not sufficient for upgrading to the CBS-V20MP2R2 model.
Recommended Action: Please abort the hardware upgrade, perform a capacity upgrade with a required step of at least 7696581394432, then start the hardware upgrade again.
Overridable: False
 
HostPathsCheck
Host  Protocol  VLAN  Paths to CT0  Paths to CT1  CT0 IOps  CT1 IOps  Status
h1	iSCSI 	- 	1         	0         	74.078	33.711	Error! No paths to CT1.
 
Message: Host(s) only have paths to one controller.
Recommended Action: Manually verify that all hosts are symmetrically connected to both controllers.
Override Warning: Since both controllers are rebooted during a Purity upgrade, ignoring this check will likely lead to an outage for the host(s) in question. Please ensure all hosts have paths to both controllers before proceeding.
Overridable: True


To proceed with the upgrade, either:
1) Retry upgrade checks - Run "purearray cloud-config model setattr --requested-model CBS-V20MP2R2 [--override <CheckName>]"
2) Abort the upgrade - Run "purearray cloud-config model setattr --requested-model ''"
3) Contact Pure Technical Services for assistance

To override the HostPathsCheck in this example, run the command:

purearray cloud-config model setattr --requested-model CBS-V20MP2R2 --override HostPathsCheck

Once the check has been overridden, the system will no longer inform the user of the Recommended Action to take for the overridden check. As you can see below, the system is only alerting the user of the CapacityStepCheck, which cannot be overridden.

purearray cloud-config model list
Current Model   	Requested Model 	Status  Step           	Details
CBS-V10MP2R2  CBS-V20MP2R2  paused  pre-upgrade-check  Upgrade has been paused: Finished pre-upgrade, 16 checks in total, 1 failed (CapacityStepCheck), 1 overridden (HostPathsCheck).
 
CapacityStepCheck
Message: Current disk capacity is not sufficient for upgrading to the CBS-V20MP2R2 model.
Recommended Action: Please abort the hardware upgrade, perform a capacity upgrade with a required step of at least 7696581394432, then start the hardware upgrade again.
Overridable: False
 
To proceed with the upgrade, either:
1) Retry upgrade checks - Run "purearray cloud-config model setattr --requested-model CBS-V20MP2R2 [--override <CheckName>]"
2) Abort the upgrade - Run "purearray cloud-config model setattr --requested-model ''"
3) Contact Pure Technical Services for assistance

In this scenario, once the capacity point is increased to the required level (V20 base or more), the CapacityStepCheck completes successfully and the upgrade process continues.

Edge Cases

An edge case can happen if, during the controller upgrade process, another issue arised and caused the upgrade to pause with error. In this situation, the array’s status is uncertain, and it may or may not have been upgraded at the time the issue arised.

Should this situation occur, we recommend contacting Everpure Technical Services for resolution.