Closed
Bug 1442369
Opened 7 years ago
Closed 7 years ago
Create a new RDS instance from latest prod snapshot to test bug 1367362
Categories
(Tree Management :: Treeherder: Infrastructure, enhancement, P1)
Tree Management
Treeherder: Infrastructure
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: emorley, Assigned: emorley)
References
Details
For bug 1367362 we need bug classification data on prototype, that only exists in production.
Jake/Kendall - please can the `treeherder-dev` RDS instance (used by the Heroku app treeherder-prototype) be reset to the most recent `treeherder-prod` RDS instance snapshot?
Since we don't yet have bug 1321294 a snapshot will need to be selected manually.
No timing coordination required; I can fix up the credentials etc afterwards (short term breakage of the dev instance is fine).
Many thanks!
Flags: needinfo?(klibby)
Flags: needinfo?(jwatkins)
Assignee | ||
Comment 1•7 years ago
|
||
I ended up creating a new temporary RDS instance to unblock bug 1367362's demo today (our Outreachy candidate's last day is today). I've added a PR to bug 1321294 with a longer term solution for resetting - at some point after that lands I'll open a new bug to actually do the reset of treeherder-dev.
I'll leave this bug open to track the cleanup of the new temporary RDS instance once bug 1367362 is merged (should a few days to a week).
Assignee: nobody → emorley
Status: NEW → ASSIGNED
Flags: needinfo?(klibby)
Flags: needinfo?(jwatkins)
Summary: Reset treeherder prototype's RDS instance to latest prod snapshot → Create a new RDS instance from latest prod snapshot to test bug 1367362
Comment 2•7 years ago
|
||
I had missed the previous update before running with the bug. Ed says that's ok, though.
Process is pretty easy now:
$ terraform taint aws_db_instance.treeherder-dev-rds
$ terraform plan
$ terraform apply
The rds instance is deleted here. since this is where I suddenly saw #c1, I control-C'ed the process. Subsequently, I had to re-run the plan/apply to create the new instance. I'm not sure if that would normally be necessary or not, but running it twice is still easier than generating a PR.
For timing, it takes ~4 minutes to destroy the RDS instance, and ~15 minutes to create from a snapshot.
Also worth noting that the new instance comes up with the password from the snapshot and should be changed manually.
Comment 3•7 years ago
|
||
Failed to add - treeherder-dev is reset.
Assignee | ||
Comment 4•7 years ago
|
||
Many thanks for doing that - I've continued the conversation about the reset in bug 1321294.
Re this (morphed) bug - I've now pointed the prototype Heroku app back at `treeherder-dev` RDS, and have deleted the `treeherder-dev-bug1367362` instance.
Status: ASSIGNED → RESOLVED
Closed: 7 years ago
Resolution: --- → FIXED
You need to log in
before you can comment on or make changes to this bug.
Description
•