Closed Bug 938263 Opened 11 years ago Closed 11 years ago

Failover git.mozilla.org to phx1 and back to scl3

Categories

(Developer Services :: General, task)

x86
macOS
task
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: fubar, Assigned: fubar)

References

Details

Tracking bug - Fail git.m.o over to PHX1 for duration of SCL3 outage, and then back again.
I assume the fail over setup will be r/o during tree closure. If not, I'm interested in how replication and fail over occurs, as we have a number of processes that automate writing.
I'm waiting to hear back from :fox2mike about whether or not PHX1 will be RW or RO. The plan thus far is:

- global RO on git.m.o and git1.dmz.phx1
- final rsync to phx1 (a couple minutes, I think; will have real numbers today)
- flip DNS (which, unfortunately, will take ~10 minutes)
- enable RW on git1.dmz.phx1

Rollback will be the same method. Of note, rsync --delete is use to ensure any files removed during failover are removed back in SCL3.
git.m.o failed over to PHX1 and returned to SCL3.
Status: NEW → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Component: WebOps: Source Control → General
Product: Infrastructure & Operations → Developer Services
You need to log in before you can comment on or make changes to this bug.