TOPIC:

i cannot upload the small .txt of 256 kb?

where can i send it?

You can sent it to This email address is being protected from spambots. You need JavaScript enabled to view it.

Thanks for providing the log file.

It looks like your Master is still running during the HA event and the replication interface is still active. Are you just pulling the management link from the master and not simulating a power-off/failure?

Since the original master is still live - the slave (new master) is unable to promote the DRBD resource to primary since the primary role is being held by the original master which is actively connected to the slave via the replication interface

The above is the root cause. TGT is getting started on top of a DRBD resource which failed to promote to primary. We'll look into possibly capturing this scenario. Please provide the exact test procedure. Specifically how the Master failure is simulated.

I didn't simulate a failed hosts. I just tried to change the roles.

1. In XenCenter: Migrate the VM to the slave host.
2. In XenCenter: Put Master in maintenance mode. Following the wizard to nominate the new master.

I See then how drbd changes the role (iscsi-cfg status)
As i saw how drbd reacts on the xen role change, i've assumed this will work. Shouldn't it?

Because its only working sometimes and when i tested it it worked everytime, i've assumed this should working. By now im facing similiar problems in the nested env. So this is expected? It's not so easy to understand what should work and what not.

If this is not possible how should I connect to XenCenter? If possible i don't want to use the cli for maintenance work.

I'll try now to manually change iscsi roles and then changing the role in xencenter as well.

Same Error after do it manually:

First i migrated the vm to xen2

1. Entered manual mode on both hosts
2. Entered secondary on xen1
3. Entered primary on xen2
4. XenCenter: Entered maintenance mode for xen1, set xen2 as new master, and the vm hangs its still pingable. i can connect to ssh, but it hangs after entering the user.

The VM came back, when I:

1. set xen2 to secondary again
2. set xen1 to primary
3. disabled manual mode on both hosts.

On the HP Installation i tried it as well with manual mode and put the master into maintenance this is what the vm shows after. The VM is still running:

end_request: I/O error, dev xvda, sector 5633016
end_request: I/O error, dev xvda, sector 5633024
Aborting journal on device dm-0-8.
EXT4-fs error (device dm-0) in ext4_reserve_inode_write: Journal has aborted
EXT4-fs error (device dm-0) in ext4_reserve_inode_write: Journal has aborted
EXT4-fs error (device dm-0) in ext4_reserve_inode_write: Journal has aborted
EXT4-fs error (device dm-0): ext4_journal_start_sb: Detected aborted journal

I will test this with a windows vm. But again it works sometimes sometimes not.

As i did everything like in the video explained. I do it manually like explained in the iscsi-ha_1.5final doc. Its still not working.

I appreciate your work and the piece of software, but i guess its not enough for what we need. This must be 100% working.