Raid failed on one DSS node

We have 2 DSS nodes with ISCSI Failover.

Now we have a problem on one DSS node. We had one disk broken and we have replaced that one. The raid and the volume was building so everything was fine. unfortunately an other disk is now also broken. The buidling process was not finished so we have a RAID Set in a building state en a volume with failed status.

Everything is swap over to the passive node and everthing is working. But the primairy is now not working. What kan i do? I have replaced the disk en the volume where DDS software is installed is back to normal but the other volume not. Stil in failing status. The RAID SET is stil in building states but he is not building. De volumes are also not available any more in DSS.

What can i do. I have nothing done else except replaced the failed one.

Regards,

Michael

You will need to let the RAID Array to complete the rebuild then once completed make sure that the DSS is running ok and the RAID health is in good standing order before moving the resources back to the local node if you are using DSS V7 if V6 then use the Sync button in the Failover service and then failback after volumes are in sync.

Hello Todd,

The problem is that the RAID SET is not building. Aretec raidcontroller said that the RAD SET is buidling but that is not true. He does nothing. The reason is that 1 one drive was broken en that one was replaced en the RAID SET was buidling. The RAID SET was not finished when the second failed. After that the VOLUME has status failed en the RAID SET still buliding. DSS is working normal but has no VOLUME because the VOLUME has the failed status.

Maybe that i must delete the RAID SET en Make a new RAID SET en make the VOLUMES. But then. After that DSS is also gone. I have a backup a the configuration en setup.

What is the best steps?

recreate a good raid, and recreate your volume groups and logical volumes. Then restore your configuration and settings, and see if you can start the cluster so you can sync back to original primary node.

I have delete the Volume en recreate with no init option. After that i see the Volumes in DSS But DSS said that there is no SYSTEM Volume.
I have repaired the volume in DSS and have no errors anymore. The next step is to sync the data from the failover server. The data on the failed DSS is not correct so i want to resync all the data. How can i do that. The Failover services is not running on the failed dss. At this moment i can't see replication task. On the failover server i see the replication tasks. The volumes are source selected.

Is the next to start the failover services. Set the volumes on destination on the failed DSS en clear the metadata on the failover server (Where the volumes has status sources). Or is there still a problem because at this moment i can't see any replication task on the failed DSS. On the failover server i see the replication task as reverse.

restore the old settings and configuration to see if the tasks come back.

Thanks for the answers. I wil do a restore Setup en configuration and hope that i got the replication task back.

The backup units holds now the Volume and is the source one.
De failed DSS (primary) is now destination. De data on the primary (failed DSS status destination) is not good so i must do a complete sync.
Where must i clear the metadata. On the backup unit where the volume is source or must i clear it on the failed one (primary) where the volume is destination.

On the failed Primary DSS if you clear the metadata in the Volume Replication keep in mind that this will do a full sync, again this is in the volume replication part of the GUI.

I have done a restore of the setup and configuration. De replication tasks are back.
I also cleared the metadata on the primary (failed DSS) en select Volume destination.

The last steps (correct me if i'am wrong) is: starting replication services on the primary DSS and select the sync button in the backup unit (wich hold the souce volume)
When the sync is completed i can do a fallback so that de primary is the active node en de backup DSS is the passive node.

Is that correct?

Thanks for the answers.

Try to start failover service from Primary and then click sync from secondary. Once synced, try failback

I have selected the sync button and that is working. The speed is about 16000 kb/s wich is not so high. It takes 50h to complete.
But when the sync is starting the performance is realy bad. We have several terminal servers and when you login it takes 10 minutes to get in. Also the linux servers get a kernel panic. After the kernel panic i reboot the linux server but then they will not boot or it takes to long. I have shutdown the primary host and everything is working normal. I dont know what i can do now. Also i dont know how i can stop the sync. I have now shutdown the primary host otherwise he is starting the sync.

How can i fix this?

You can stop the Sync from the Volume Replication task, the sync needs to complete, make sure that the volume replication tasks is consistent to consistent for both Source and Destination in Status > Tasks, this is most likely the issue of your performance.