Closed Bug 1413542 Opened 7 years ago Closed 5 years ago

Update Treeherder RDS instances from MySQL 5.7.17 to 5.7.23

Categories

(Tree Management :: Treeherder: Infrastructure, enhancement, P2)

enhancement

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: emorley, Assigned: emorley)

Details

Attachments

(5 files)

Assignee: nobody → emorley
Status: NEW → ASSIGNED
Attachment #8952087 - Flags: review?(jwatkins)
Comment on attachment 8952087 [details] [review]
devservices-aws PR #79: Upgrade treeherder-dev MySQL

I'm going to 302 to :fubar since I'm swamped with datacenter move deadlines
Attachment #8952087 - Flags: review?(jwatkins) → review?(klibby)
Comment on attachment 8952087 [details] [review]
devservices-aws PR #79: Upgrade treeherder-dev MySQL

Currently applying update; will update when it's finished.
Attachment #8952087 - Flags: review?(klibby) → review+
Done.
Status: ASSIGNED → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
Many thanks - will leave this bug open to track verifying no issues with 5.7.19 and then updating stage/prod to use it too.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Upgrade log looks fine:
https://emorley.pastebin.mozilla.org/9079148
5.7.21 is now available, so lets target that instead:
https://dev.mysql.com/doc/relnotes/mysql/5.7/en/news-5-7-20.html
https://dev.mysql.com/doc/relnotes/mysql/5.7/en/news-5-7-21.html
Summary: Update Treeherder RDS instances from MySQL 5.7.17 to 5.7.19 → Update Treeherder RDS instances from MySQL 5.7.17 to 5.7.21
Hi! This can be applied at any time - no need for coordination.
Later I'll open another PR for stage/prod (so the bug should stay open).
Many thanks :-)
Attachment #8967358 - Flags: review?(klibby)
Status: REOPENED → ASSIGNED
Comment on attachment 8967358 [details] [diff] [review]
devservices-aws PR #83: Upgrade treeherder-dev MySQL again

Review of attachment 8967358 [details] [diff] [review]:
-----------------------------------------------------------------

approved, merged, about to deploy...
Attachment #8967358 - Attachment is patch: true
Attachment #8967358 - Attachment mime type: text/x-github-pull-request → text/plain
Attachment #8967358 - Flags: review?(klibby)
Attachment #8967358 - Flags: review+
Attachment #8967358 - Flags: checkin+
aws_db_instance.treeherder-dev-rds: Modifying... (ID: treeherder-dev)
  engine_version: "5.7.19" => "5.7.21"
aws_db_instance.treeherder-dev-rds: Still modifying... (ID: treeherder-dev, 10s elapsed)
aws_db_instance.treeherder-dev-rds: Still modifying... (ID: treeherder-dev, 20s elapsed)
aws_db_instance.treeherder-dev-rds: Still modifying... (ID: treeherder-dev, 30s elapsed)
aws_db_instance.treeherder-dev-rds: Modifications complete after 32s (ID: treeherder-dev)

and I always forget apply_immediately, so manually forced the upgrade via console. and done an hour later, due to making two db backups/snapshots.
Oops meant to come back to this sooner. Now 5.7.23 is available:
https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/CHAP_MySQL.html#MySQL.Concepts.VersionMgmt

https://dev.mysql.com/doc/relnotes/mysql/5.7/en/news-5-7-22.html
https://dev.mysql.com/doc/relnotes/mysql/5.7/en/news-5-7-23.html

I'll open a PR to update dev to 5.7.23 and follow up with the others soon after.
When you apply, could you use apply_immediately? :-)
Summary: Update Treeherder RDS instances from MySQL 5.7.17 to 5.7.21 → Update Treeherder RDS instances from MySQL 5.7.17 to 5.7.23
Attachment #9030673 - Flags: review?(klibby) → checkin+

Please could you apply this when you get a chance (using apply immediately again).
I believe stage should take less time than the previous one, since it's multi-AZ so the upgrade fails over instead of taking it offline.

Attachment #9035280 - Flags: review?(jwatkins)
Attachment #9035280 - Flags: review?(jwatkins) → review+

(In reply to Ed Morley [:emorley] from comment #14)

Created attachment 9035280 [details] [review]
devservices-aws PR #95: Upgrade treeherder-{stage,prod-ro} MySQL to 5.7.23

Please could you apply this when you get a chance (using apply immediately again).
I believe stage should take less time than the previous one, since it's multi-AZ so the upgrade fails over instead of taking it offline.

Merged, applied and upgrades completed.
aws_db_instance.treeherder-prod-ro-rds: Modifications complete after 16m18s (ID: treeherder-prod-ro)
aws_db_instance.treeherder-stage-rds: Modifications complete after 56m49s (ID: treeherder-stage)

Thank you :-)

Hi Jake!

This applies the same change to prod as was applied to dev/stage/prod-ro.

When it was applied for stage it caused 2-5 mins of errors as reported by New Relic which are for tasks that retry, so it shouldn't cause significant impact. However it may be best to apply this at at lower load time of day (eg early pacific) to reduce that further.

Many thanks :-)

Attachment #9040971 - Flags: review?(jwatkins)
Comment on attachment 9040971 [details] [review]
devservices-aws PR #96: Upgrade treeherder-prod MySQL to 5.7.23

I'll apply this on Monday 2019/02/11 @ 0900 PST
Attachment #9040971 - Flags: review?(jwatkins) → review+

That sounds great - thank you :-)

Comment on attachment 9040971 [details] [review]
devservices-aws PR #96: Upgrade treeherder-prod MySQL to 5.7.23

aws_db_instance.treeherder-prod-rds: Modifications complete after 1h3m21s (ID: treeherder-prod)
Attachment #9040971 - Flags: checkin+

Many thanks!

Terraform change landed in:
https://github.com/mozilla-platform-ops/devservices-aws/commit/6db0b6441529dba4a38c023813a6f003b5cf779f

Looking at the AWS RDS event logs, the DB started its restart at 17:38:17 and was back at 17:41:43, so much quicker than stage.

Status: ASSIGNED → RESOLVED
Closed: 6 years ago5 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: