Metadata service monitoring
Bug #2034687 reported by
Nishant Dash
This bug affects 2 people
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| OpenStack Nova Compute Charm |
Triaged
|
Wishlist
|
Unassigned | ||
| charm-ovn-chassis |
New
|
Undecided
|
Unassigned | ||
Bug Description
There needs to be a method to be able to monitor connectivity to the metadata server 169.254.169.254
I am not exactly sure where to place this request, but at its core nova should be able to tell us if any nova-api-metadata service on a host is not functioning as expected.
A recent occurrence I have seen is,
1) Cluster suffers from networking outage
2) We recover a bunch of services like ceph, nova compute service
3) days later vm creations have issues connecting to metadata server
4) check nova-api-metadata service logs and it initially reported some AMQP timeouts days ago and then STOPPED logging entirely after that, but the systemd service is running "green"
| description: | updated |
| description: | updated |
| tags: | added: canonical-is |
| Changed in charm-nova-compute: | |
| importance: | Undecided → Wishlist |
| status: | New → Triaged |
Revision history for this message
| Felipe Reyes (freyes) wrote | #1 |
To post a comment you must log in.
Adding charm-ovn-chassis for ovn-based deployments that run neutron-