RecoverPoint for VMs setup issue

Hi,

Can you check if the selected datastore to contain the vRepository can be accessed through all ESXs housing the vRPAs ?

If the datastore cannot be accessed by all RPAs, this error might be issued.

Regards,

Idan

Yes, it can.  This is a standalone ESXi server using local storage.

Please make sure that the vCenter credentials you used in the previous Deployment Manager step are of a vCenter administrator role. Some roles cannot create VMDKs which might fail the step you're in because the vRepository VMDK should be created on the selected datastore.

Also, before retrying, make sure to remove the RPvStorage folder from the datastore you've attempted to create the vRepository on previously.

Hope that helps,

Idan

I have validated that the vCenter credentials used were configured with the vCenter administrator role.  Anything else I can try?

Yes, which RP user have you entered in the vCenter Server Registration step ?

It is a full admin in the vCenter environment.

I was referring to the RecoverPoint user used in Deployment Manager as part of the vCenter Registration step.

Is it possible to for you to attach the logs and oldlogs directories of Deployment Manager ? I would like to take a close look at the issue you're experiencing.

Regards,

Idan

Can confirm the paths of the directories you need sent over?

Sure, both directories are under the deployment manager directory:

../RecoverPoint_rel4.2_mercury.92_DM_2.2_Win_md5_c816030b948621c09336cd6c37fbe032/logs

../RecoverPoint_rel4.2_mercury.92_DM_2.2_Win_md5_c816030b948621c09336cd6c37fbe032/oldlogs

Idan

https://accudatasystems.sharefile.com/d/sf71673394e84928a

The logs can be at the link above

Thanks.

Based on our conversation, please try with RP admin account during the VC registration step and let me know.

I ran into the same issue.  I was able to resolve it by verifying the following three things (these were also mentioned above):

1) Ensure the datastore selected to contain the vRepository can be accessed through all ESXs housing the vRPAs

2) Ensure that you are using the RP admin account during the VC registration step.  The defaults are shown in the RecoverPoint for VMs 4.2 Administrators Guide.

3) Ensure that the RPvStorage folder is removed from the datastore you've attempted to create the vRepository on previously.

Once I performed all three of these steps and retried I was able to get past the "Internal Error"

I hope this helps.

All accurate points, another known issue is when receiving an internal error on scaled environments, the creation of the Repository VMDK might timeout. The workaround is to increase that timeout in Deployment Manager. This is also mentioned in the RecoverPoint Uptime Bulletin - Q4'14 along with the link to relevant KB.

Regards,

Idan

Hello Idan,

I am facing exactly the same issue and the dm.log contains the following error:

2015-02-03 14:39:59,856 [pool-1-thread-39] (BaseInstallationServerAdapter.java:336) ERROR - Transaction failed. transactionId=67, timeoutInSeconds=3120, errorMSG=Unable to execute command. . Internal error. Errors encountered when writing to repository volume, please make sure the volume is not read only., errorType=UNABLE_TO_EXECUTE_COMMAND, value=null

2015-02-03 14:39:59,856 [pool-1-thread-39] (ConfigureRepositoryOnSpecificRPACommand.java:88) DEBUG - repository format finished.

2015-02-03 14:39:59,856 [pool-1-thread-39] (BaseDMCommand.java:150) ERROR - ConfigureRepositoryOnSpecificRPACommand FAILED

The datastore i select is an FC vmfs datastore that is visible from both esxi hosts in the cluster.

Hi there, indeed this known issue is documented in KBA196220.

Let me know if you cannot access this article.

Regards,

Idan