Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

masakari notification on process failure

Hi all, I have setup a multi host set up using devstack stein with 1 controller node and 2 compute nodes with ubuntu 16 VMs. I have enabled Masakari service on my controller local.conf file like below -

#Enable Masakari
enable_plugin masakari ${GIT_BASE}/openstack/masakari $USE_BRANCH
#enable_plugin masakari-dashboard ${GIT_BASE}/openstack/masakari-dashboard $USE_BRANCH
enable_plugin masakari-monitors ${GIT_BASE}/openstack/masakari-monitors $USE_BRANCH
enable_plugin python-masakariclient ${GIT_BASE}/openstack/python-masakariclient $USE_BRANCH

I have set up Masakari Segments adding 2 compute nodes in it with AUTO recovery under the 'Instance-ha' section in horizon dashboard. I have to mimic process, VM and compute failure to check whether Masakari provides proper notification in each failure case.

To test process failure, I have created 2 instances in one compute. From root user, I killed the instance 1's qemu process on that particular host. However masakari did not provide notification for this process failure if it is monitoring the underlying processes on controller and computes.

Please let me know if I am missing anything in configuring Masakari and how Masakari notifications really works to provide useful information to the end user.

thank you for your pointers and replies.

best regards,

masakari notification on process failure

Hi all, I have setup a multi host set up using devstack stein with 1 controller node and 2 compute nodes with ubuntu 16 VMs. I have enabled Masakari service on my controller local.conf file like below -

#Enable Masakari
enable_plugin masakari ${GIT_BASE}/openstack/masakari $USE_BRANCH
#enable_plugin masakari-dashboard ${GIT_BASE}/openstack/masakari-dashboard $USE_BRANCH
enable_plugin masakari-monitors ${GIT_BASE}/openstack/masakari-monitors $USE_BRANCH
enable_plugin python-masakariclient ${GIT_BASE}/openstack/python-masakariclient $USE_BRANCH

I have set up Masakari Segments adding 2 compute nodes in it with AUTO recovery under the 'Instance-ha' section in horizon dashboard. I have to mimic process, VM and compute failure to check whether Masakari provides proper notification in each failure case.

To test process failure, I have created 2 instances in one compute. From root user, I killed the instance 1's qemu process on that particular host. However masakari did not provide notification for this process failure in horizon instance-ha notifications if it is monitoring the underlying processes on controller and computes.

Please let me know if I am missing anything in configuring Masakari and how Masakari notifications really works to provide useful information to the end user.

thank you for your pointers and replies.

best regards,