Hi folks,
I try to get drbd9 working on my servers. My configuration is following:
I have 5 nodes, 2 of them are primary fileserver (fs1 and fs2)
3 are virtualisation hosts running proxmox (virt1…virt3)
All of them have network cards for derby connections:
10.10.10.33/26
The virtualisation hosts have additional network cards.
Running newest drbd from linbit repo.
After several hours all nodes seem to be connected and talk to each other.
(drbdmanage n —> all have OK state)
+------------------------------------------------------------------------------------------------------------+
| Name | Pool Size | Pool Free |
| State |
|------------------------------------------------------------------------------------------------------------|
| fs1 | 7630888 | 6479712 |
| ok |
| fs2 | 7630888 | 6436592 |
| ok |
| virt1 | 19260 | 19252 |
| ok |
| virt2 | 19260 | 19252 |
| ok |
| virt3 | 19260 | 19252 |
| ok |
+------------------------------------------------------------------------------------------------------------+
With drbdadm I can see that fs1 is primary control node.
When trying to deploy a VM or move storage to drbd I nearly always get "TASK
ERROR: storage migration failed: drbd error: Could not forward data to leader"
Sometimes this is working for the setting of "redundancy 1"; the setting of
"redundancy 2" or higher never worked. In higher redundancy I can see in log
file something like "Initial split brain detected" and it fails completely.
After the storage correctly runs on redundancy-1-level, I can assign it to the
second fileserver.
When doing this while the deploy is running it will most likely fail at 99.8%
or so with some blocks not syncing.
So what am I possibly doing wrong?
Thanks in advance for useful hints,
cheers, Frank
_______________________________________________
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user
