We're updating the issue view to help you get more done. 

Redolog metrics stops after restarting the primary

Description

Reproduction:
1. config metrics
2. start the grid with one primary and one backup
3. deploy space with EDS (attached module that you can use).
4. do not deploy mirror.
5. start benchmark.
6. as the mirror is down, you will see the redolog raising in the ui
7. in your metric reporter you will see that the value of space_replication_redo-log_size is also raising (this is the primary redolog saved for the mirror)
8. restart the primary.

Expected:
the backup become primary and his space_replication_redo-log_size metric will keep raising.

Actual result:
from this point the primary that become backup and got a new pid starts report on space_replication_redo-log_size = 0 and the backup that become primary doesn't report at all on this metric.

Seems that we need to do new registration "after primary"

Workaround

None

Acceptance Test

ReplicationMetricWithBackupAndMirrorDownTest none none
ReplicationMetricWithBackupTest none none

Status

Assignee

Yohana Khoury

Reporter

Yuval Dori

Labels

None

Priority

Medium

SalesForce Case ID

10516, 11883

Fix versions

Commitment Version/s

None

Due date

None

Product

XAP

Edition

Open Source

Platform

All