Closed
Bug 685746
Opened 14 years ago
Closed 14 years ago
[amo] database dumps on cm-webdev01-master01 are stale
Categories
(Infrastructure & Operations Graveyard :: WebOps: Other, task)
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: clouserw, Assigned: bhourigan)
Details
cm-webdev01-master01:/data/backup-drop/mrdb04/mysql/addons_remora
We're still getting daily dumps there which is why we're just now noticing this, but judging by file sizes the dumps aren't changing, just doing the same thing over and over. Last useful dump was from Aug 24th. We should be getting dumps from the db in phoenix.
| Reporter | ||
Comment 1•14 years ago
|
||
This is causing us much pain in bug 685669
Comment 2•14 years ago
|
||
As you may have guessed, this is broken due to the PHX1 migrations.
The dump that gets placed there is created on tm-backup02, which is a slave of tm-amo-master02. This is one of the original DB masters in SJC. Of course, the data there is no longer being updated, so the dumps don't show any new data.
Dumps on tm-backup02 are all rsync'd to dracula (this is not AMO-specific... all backup dumps do this). Dracula, has a number of cron jobs to copy dumps out to wherever they're needed. This is how the dumps ultimately end up on cm-webdev01-master01.
There is a project underway to set up a DB backup server in PHX1. Once that's done the dumps would originate from there, instead of tm-backup02... this is the longer-term fix, and should be done in the next week or two.
In the interim, we can try to do this manually... that is, make a dump of the data from PHX1 off of a slave, and put that on cm-webdev01-master01 somewhere (and probably disable the cron job that automatically updates the database with the most current copy from tm-backup02, which would be counter-productive at that time).
Whether or not this would be useful to you is hard to judge... the dumps would be similar to the ones discussed in the bug linked in comment 1, which I understand has already been provided to you.
Comment 3•14 years ago
|
||
(In reply to Jake Maul [:jakem] from comment #2)
> There is a project underway to set up a DB backup server in PHX1. Once
> that's done the dumps would originate from there, instead of tm-backup02...
> this is the longer-term fix, and should be done in the next week or two.
whoa wait, this should already be done and backing up AMO dumps... Dumitru? Phong?
Comment 4•14 years ago
|
||
(In reply to Corey Shields [:cshields] from comment #3)
> whoa wait, this should already be done and backing up AMO dumps...
> Dumitru? Phong?
The backup part of the data is already done. See Bug 682284. Not sure about the status of the sql dumps themselves, mpressman can shed some light here.
Comment 5•14 years ago
|
||
We have a more-recent-than-8/24 dump in there right now but I believe the path of getting automatic updates is still broken.
Brian can you work with dumitru on getting the dumps from their backup source into cm-webdev01-master01 in sjc1? (the dumps are from prod in phx1)
Assignee: server-ops → bhourigan
| Assignee | ||
Comment 6•14 years ago
|
||
Awaiting Dumitru's return from PTO to work with him on this.
Updated•14 years ago
|
Assignee: bhourigan → server-ops
Component: Server Operations → Server Operations: Web Operations
QA Contact: mrz → cshields
Updated•14 years ago
|
Assignee: server-ops → bhourigan
| Assignee | ||
Comment 7•14 years ago
|
||
Bug 691590 created to allow network flow from db-backup1 to cm-webdev01-master01. An initial dump from today is being loaded into cm-webdev01-master01.remora and the script to pull it from tm-backup02 has been temporarily disabled.
| Assignee | ||
Comment 8•14 years ago
|
||
Data will regularly be dumped to cm-webdev01-master01 from db-backup1 into /data/backup-drop/addons/mysql/addons_mozilla_org nightly at 2AM
Status: NEW → RESOLVED
Closed: 14 years ago
Resolution: --- → FIXED
| Reporter | ||
Comment 9•14 years ago
|
||
There's a file from yesterday at 2130 that is empty. Nothing from 2am.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
| Assignee | ||
Comment 10•14 years ago
|
||
Sorry about that, there was a bug in the script. There is a 3.8GB dump file present now.
Status: REOPENED → RESOLVED
Closed: 14 years ago → 14 years ago
Resolution: --- → FIXED
| Reporter | ||
Comment 11•14 years ago
|
||
There is a dump from yesterday (thanks!) but nothing from 2am today
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
| Assignee | ||
Comment 12•14 years ago
|
||
Ack! I'm really sorry about the trouble, Wil. I've made a change and I'm creating a new dump for you now. Let's see if it works tonight at 2AM as intended.
| Assignee | ||
Comment 13•14 years ago
|
||
It looks like it ran last night, please confirm.
| Reporter | ||
Comment 14•14 years ago
|
||
Looks good to me. Thanks!
| Assignee | ||
Updated•14 years ago
|
Status: REOPENED → RESOLVED
Closed: 14 years ago → 14 years ago
Resolution: --- → FIXED
Updated•12 years ago
|
Component: Server Operations: Web Operations → WebOps: Other
Product: mozilla.org → Infrastructure & Operations
Updated•7 years ago
|
Product: Infrastructure & Operations → Infrastructure & Operations Graveyard
You need to log in
before you can comment on or make changes to this bug.
Description
•