We should increase time to live for messages and queues for max value in mcollective
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Fix Released
|
High
|
Vladimir Sharshov | ||
4.1.x |
Fix Released
|
High
|
Fuel Library (Deprecated) |
Bug Description
{"build_id": "2014-05-
Steps to Reproduce:
1. Deploy one node cluster using KVM
2. Delete it
3. Deploy neutron gre cluser with 4 nodes (1controller 2 computes 1 cinder+ mongo) with murano/
Expected
Deployment pass , cluster ready, ostf pass
Actual:
deployment of controller fail with message [7ff473ade700] (receiver) Node slave-01_controller not answered by RPC, removing from db
Info: Dima P comment: need to set ttl to max value for mcollective
no longer affects: | qemu-kvm (Ubuntu) |
Changed in fuel: | |
importance: | Critical → High |
Changed in fuel: | |
assignee: | Fuel Python Team (fuel-python) → Andrey Danin (gcon-monolake) |
Changed in fuel: | |
assignee: | Andrey Danin (gcon-monolake) → Vladimir Sharshov (vsharshov) |
Changed in fuel: | |
status: | Fix Committed → Confirmed |
we have failed tests on CI especially for ha with a very similar problem. We marked node with error status with 2014-05-06T14:56:36 err: [394] MCollective agents '1' didn't respond within the allotted time.
and do not wait a little bit more , but than we can see in logs that node is responsible and online