behaviour of innodb_io_capacity contradicts documentation
Affects | Status | Importance | Assigned to | Milestone | ||
---|---|---|---|---|---|---|
Percona Server moved to https://jira.percona.com/projects/PS | Status tracked in 5.7 | |||||
5.1 |
Won't Fix
|
Medium
|
Unassigned | |||
5.5 |
Triaged
|
Medium
|
Borys Belinsky | |||
5.6 |
Invalid
|
Undecided
|
Unassigned | |||
5.7 |
Invalid
|
Undecided
|
Unassigned |
Bug Description
Version: percona-server 5.1.56 with innodb 1.0.15-12.5
Description:
According to documentation linked below, and an informal chat with Vadim, there is no need to tune innodb_io_capacity when innodb_
How to reproduce:
* on replication slave with FusionIO storage, set
innodb_
innodb_
* then add significant write load to master
* observe innodb checkpoint age stay near max age
* then run mysql> SET GLOBAL INNODB_
* observe that Innodb_
Suggested work-around:
Simply tune innodb_io_capacity.
summary: |
- adaptive_checkpoint = keep_average requires tuning of io_capacity + behaviour of innodb_io_capacity contradicts documentation |
Changed in percona-server: | |
status: | New → Triaged |
importance: | Undecided → Medium |
tags: | added: doc |
Changed in percona-server: | |
status: | Triaged → In Progress |
Changed in percona-server: | |
assignee: | nobody → Hrvoje Matijakovic (hrvojem) |
Changed in percona-server: | |
status: | In Progress → Confirmed |
Changed in percona-server: | |
status: | Confirmed → Triaged |
I tested this behaviour on the same version of percona-server with innodb_ adaptive_ checkpoint = estimate, and on hardware with a traditional HDD RAID. In this case, tuning innodb_io_capacity still has an immediate and significant impact on flush rate, which directly contradicts the docs. The docs state that for estimate, "behavior is independent of the innodb_io_capacity variable".