TOPIC:

Hi,

Am trying out the XenServer 7 beta. I am doing testing, and am attempting to make the Pool Master Storage Secondary and the Slave Storage Primary (to test how I would perform maintenance tasks on the hosts).

I have an error after the actions below:

1.) Turned off ha (ha-cfg status then "yes" to disable)
2.) Put both servers into manual mode (iscsi-cfg manual-mode-enable)
3.) Moved all VM's to slave
4.) demote Master storage (iscsi-cfg become-secondary)

At this point we get a pause for about 10 seconds and then I get the following output:

[root@COR-SVR-C01214 ~]# iscsi-cfg become-secondary
/bin/iscsi-cfg: line 274: 11990 Terminated /etc/iscsi-ha/iscsi-ha.sh become_secondary
###########################################################
# iscsi-ha is in manual mode - current status shown below #
###########################################################
Storage role: Primary (Expected = Secondary)
Replication IP: 10.10.10.1/24
iSCSI target: deactivating (stop-sigterm) (Result: exit-code) since Mon 2016-10-10 12:09:08 BST; 32s ago (Expected = stopped)
/bin/iscsi-cfg: line 31: /etc/iscsi-ha/state/status: Permission denied
[root@COR-SVR-C01214 ~]#


Any help much appreciated.

Regards,

Adam

Just to note, it would appear if I try this two or three times and then wait a few minutes, the storage does become secondary...

A bug maybe? Also, when you then try and reboot the server that you've set to secondary storage, it hangs at the XenServer unloading splash screen (see attached). You have to force a reboot (via iLO - power Reset) to bring the server back online.

Finally - if you disable HA, turn off all VM's and then shut down the Slave server, the Slave server shuts down cleanly. If you then shut down the Master, it hangs at the XenServer Shutting down splash screen (as per screenshot in previous post). You can only get the server working again by a power reset in iLO.

Regards,

Adam

the tgt daemon requires a system drop-in file in order to shutdown the daemon rapidly.

Check for the existence of:
/etc/systemd/system/tgtd.service.d/local.conf

the contents should be
[Service]
TimeoutStopSec=2

You can create the file manually if it is not there.

Salvatore Costantino wrote: the tgt daemon requires a system drop-in file in order to shutdown the daemon rapidly.

Check for the existence of:
/etc/systemd/system/tgtd.service.d/local.conf

the contents should be
[Service]
TimeoutStopSec=2

You can create the file manually if it is not there.

I will give this a try in a couple of hours when I'm back at work and post back the outcome.

Thanks!

Adam

Adam Ward wrote:

Salvatore Costantino wrote: the tgt daemon requires a system drop-in file in order to shutdown the daemon rapidly.

Check for the existence of:
/etc/systemd/system/tgtd.service.d/local.conf

the contents should be
[Service]
TimeoutStopSec=2

You can create the file manually if it is not there.

I will give this a try in a couple of hours when I'm back at work and post back the outcome.

Thanks!

Adam

Hi,

I can confirm that the local.conf file in /etc/systemd/system/tgtd.service.d already exists on both servers and the contents are:

[Service]
TimeoutStopSec=2

Any ideas on why the server running as the primary iSCSI storage fails to shutdown and what the error message posted above are?

Happy to provide any logs.

Regards,

Adam

Excuse the late reply. We'll give this a try and let you know if it can be reproduced.