Clone Controlled Replication issues in 9.1.1.1

Question

We have upgraded our networker version from version 9.0.1.6 to version 9.1.1.1, because we are upgrading our vSphere environment to version 6.5 and want to make use of the new vproxies. After upgrading to this version we are having problems using cloning between our two data domains. The backups work fine and the clones to tape also work fine. The clone from one data domain to another is behaving erratic. Mostly clones after the backup of our networker clients starts to hang when another clone starts that wants to access the same volumes. (e.g. NAS backups). When we kill the clone of the clients, the other clone starts immediately and finishes without a problem. In the logs I find some references to server parallelism that is not set high enough, but is currently set at 1024 which is the maximum value. In the logs of the clone I find some warnings that weren't there is the previous version of networker, maybe this has got something to do with it? these warnings are: - Unable to join save thread during clone end. The thread is NULL - Unable to join recovery thread during rs context clean up . The thread is NULL - Unable to join save thread during rs context clean up . The thread is NULL Has somebody experience the behaviour that I described here. Or do know what these warnings mean? I already opened a SR, but by the time it is assigned and that the problem is really investigated, we are a few days further and I wanted to get a headstart.

nowx · Answer

Hi,

it´s an open issue.

https://emcservice.force.com/CustomersPartners/kA2j0000000fyBwCAI

You have to enable "RPS Clone" for DD->DD clones. But with RPS Clone is activated (not disabled in configiguration), Clones to Tape does´t work.

RPS enabeld -> DD-DD clones are working.

RPS disabled -> Tape clones are working.

atm, you have to decide...

br, daniel

alcreton · Answer

Hi We also have the same problem since we have upgraded to NETWORKER 9.1.0.6. DDOS Version is 6.0.1.10-561375. 6/20/2017 03:01:57 AM Cloning successfully done for save set : WINDOWS ROLES AND FEATURES:\ (ssid 977804901) 06/20/2017 03:01:57 AM Size cloned for save set: WINDOWS ROLES AND FEATURES:\ (ssid 977804901) is 6335 KB 06/20/2017 03:01:57 AM Total size cloned so far: 6335 KB.
 06/20/2017 03:01:57 AM Unable to join save thread during clone end. The thread is NULL 06/20/2017 03:01:58 AM Unable to join recovery thread during rs context clean up . The thread is NULL 06/20/2017 03:01:58 AM Unable to join save thread during rs context clean up . The thread is NULL 06/20/2017 03:02:00 AM 06/20/17 03:02:00.329935 Total time taken to replicate ssid file /mcq-bkp50/DD50ORPDA1MO_1/88/44/873e79d8-00000006-574819ef-594819ef-02ab5000-4057ae56 is 72. 06/20/2017 03:02:00 AM 06/20/17 03:02:00.330064 Total time taken to replicate ssid file /mcq-bkp50/DD50ORPDA1MO_1/88/44/873e79d8-00000006-574819ef-594819ef-02ab5000-4057ae56 is 72. Unchecking the the RPS Clone doesn't solve the problem. We are going to open a new case with EMC support team. We have a non production DATA ZONE with the NETWORKER/DDOS version and don't have any problem with this platform Any other Idea

ble1 · Answer

In my case, I have seen it with AFTD (in Azure) during nsrclone -m (so traditionally what we called staging).  Disabling RPS fixed this. Not sure yet (not tested yet) how this works with CloudBoost devices.  I know, from some other people, that there is a custom patch where you can control this on group level (which is something folks will need if and when mixing DD and tapes). Not sure what current plans are within engineering how this will be addressed further.

otmanix · Answer

I have a new Networker 9.1.0.5 installation. I'd like to update to 9.1.0.6 because of support for VMware VSphere versions.

We have 2 sites and a dd4500 at each site. I have workflows which do backups to the local dd4500 and then clone to the remote dd4550. I'm doing a migration from old 8.2 installtion and also have to configure NDMP backups to local DD with clone to remote DD.

At the moment erything works as expected. But I'm afraid that after updating new Networker installation to >9.1.0.5 that I may fix VMware problems but get trouble with clones.

Are there any news when there will be a fix?

Regards, Otmar

NetWorker

Was this post helpful?