TOPIC:

"Failed to update global pool configuration" is a warning that could be triggered if the pool momentarily has no master. Since all HA-Lizard configuration parameters are centrally stored in the XAPI DB, HA-Lizard will update a local configuration cache roughly every minute with the global parameters. HA-Lizard will wait 10 seconds to update its cache and will then trigger this alert if the timeout has been reached.

This could happen while a pool master is being recovered.

Simon,
Version 1.6.42 has been posted in the downloads section and is available for download. To upgrade your servers:

service ha-lizard stop -w

then run the installer - select "no" to install pool global params. If you have made changes to your configuration override files (ha-lizard.conf) you will need to re-make those changes in the new installation.

This version resolves the issue you reported and adds some enhancements:

- Validated support for XenServer 6.2
- Updated ha-cfg tool to warn user that a reset is required when updating monitor timers
- Updated ha-cfg tool to warn that FENCE_HOST_FORGET will be deprecated
- Improved email alert handling prevents duplicate messages from being sent for a configurable timeout - default 60 minutes
- Updated email alert message notifying user of failed or started VM. Message now includes the VM name-label AND UUID
- Improved recovery from system crash which could erase configuration cache if crash occurs while updating configuration
- FD leak caused interrupted system calls which closed TCP sockets used to check host status - Moved TCP socket connection from FD to alternate method resolving issue with email alerts warning of failure to connect to HTTP port of slaves.

I am getting these notifications as well, but I think it is due to one or more of the VMs when they are using high CPU. My hosts only have dual core procs. Is this anything to worry about, and which parameter would apply to increase this threshold of 10 seconds to something like 30 seconds? Thanks in advance.

Are you running an old version of HA-Lizard (older than 1.6.42)?

If not, this is possible if:
- your management interface does not respond to ping when under heacy CPU load. You can test this with a continuous ping while your VM is under heavy CPU load.

OR

- XAPI is not responsive when a VM is under heavy CPU load. We have seen poor XAPI performance on underpowered hosts. You can also give this a quick test by connecting with a browser to the host's management IP while it is under heavy CPU load.. If the reponse is delayed or there is no response this would be the cause of your issue.

You can probably decrease the frequency of the alerts by increasing your monitor delay to about 1 minute.
"ha-cfg set monitor_delay 60"

sc wrote: Are you running an old version of HA-Lizard (older than 1.6.42)?

I'm on 1.7.8. I will try your suggestions and post back any further issues. Thank you very much.

It seems in my case, I'm still getting emails. I thought it was due to high CPU usage on the VMs, but now I'm not so sure. I have a pool of two hosts, and occasionally they will both send an email around the same time, that looks like this:

xe_wrapper: COMMAND: xe host-param-list uuid=e9df4bb9-18f4-423b-8cf8-2d58dfcf4d33 Has reached the maximum allowable time of 10 seconds. Killing all processes now!

xe_wrapper: COMMAND: xe vm-list is-control-domain=false resident-on=e9df4bb9-18f4-423b-8cf8-2d58dfcf4d33 Has reached the maximum allowable time of 10 seconds. Killing all processes now!

What does this mean, and is there a workaround? Everything seems to be running fine as far as the VMs go on each host. Again I'm on HA-Lizard 1.7.8 on XenServer 6.2 with all of the latest patches.

Thank you for your help.