TOPIC:

I have been working on a test pool setup to troubleshoot problems on my main pool setup which I have paperwork pending for a support session. So i simulated a power failure on my master and the the pool successfully transfer responsibilities over to the slave. Upon reconnecting the former master back into the pool I have a current situation in which now that i original master back as the primary server and my slave set as the secondary server i have DRBD reporting a wfconnection state on the primary and a standalone on the secondary.

- Xencenter reports connectitivy to both my iscsi SR
- I can enter manual mode swap primary and secondary roles and reboot both servers when either is in the secondary manual mode and the vm's stay active.

it just appears that DRBD has some connectivity issue.

smlog

DRBD may have triggered split brain which prevents the resource from reconnecting. Try searching the logs around the time of the test for a DRBD split brain notice. If confirmed, you can follow the DRBD docs or ours which describe how to resolve it.

following your documentation fixed the split brain senario.

I was then able to put my original master and slave back into their proper states. I proceeded to migrate my vm's to the slave and do another round of testing starting with pulling the plug on the slave. The vm's successfully migrated and reboot on the master. I then booted the slave back up and currently I can't get the SR to plug to the slave server. DRBD, HA-lizard, and iscsi-ha are all running and not reporting issues. the drbd backend is connected and both iscsi1 resources show updated on the primary and secondary.

Here is a sample of my var/log/messages showing that while the drbd resources and connected the volume groups are not found.



Jun 8 08:45:54 ctxen-ha06 iscsi-ha: 10309 DRBD Resource: iscsi1 in Connected state
Jun 8 08:45:54 ctxen-ha06 iscsi-ha: 10309 DRBD Resource: iscsi2 in Connected state
Jun 8 08:45:54 ctxen-ha06 iscsi-ha-NOTICE-/etc/iscsi-ha/iscsi-ha.sh: No volume groups found
Jun 8 08:45:54 ctxen-ha06 iscsi-ha-10270-INFO-/etc/iscsi-ha/iscsi-ha.sh: Scanning for Volume Group -> iscsi-sr: 3a7ab9cd-6d0f-c867-31bc-758f3a4a9cd2
Jun 8 08:45:54 ctxen-ha06 iscsi-ha-10270-INFO-/etc/iscsi-ha/iscsi-ha.sh: Scanning for Volume Group -> iscsi-sr: 5455150d-aa3e-eb5a-b650-3a52c9686a86
Jun 8 08:45:56 ctxen-ha06 xapi: [ info|ctxen-ha06|279876 INET :::80|dispatch:PBD.plug D:eb5758203d7b|taskhelper] task PBD.plug R:e629bad604da forwarded (trackid=6842ed027cdf56231e6066a24c22aa1c)
Jun 8 08:45:56 ctxen-ha06 xapi: [ info|ctxen-ha06|279876 INET :::80|PBD.plug R:e629bad604da|storage_access] SR 3a7ab9cd-6d0f-c867-31bc-758f3a4a9cd2 will be implemented by /services/SM/lvmoiscsi in VM OpaqueRef:59b65636-4c4d-a659-ba44-e16523a28fbf

try "service tgtd restart" on the master and see of that fixes it.