826694 - Update boot.scr on all panda boards

Status: RESOLVED FIXED

(Infrastructure & Operations :: RelOps: General, task)

Description

[Jake Watkins [:dividehex]]

We need to update the boot.scr and boot.txt on all panda boards.  I have already created a new preseed image with this change that can be used for new panda boards and any others that need manual intervention.  For panda boards already in mozpool, i have also created a fix-boot-scr "image" that will pxe boot a panda board, update its boot.scr/txt and drop into maintenance mode.

We will be updating any board not in a locked_out state.  For those in a locked out state, I will coordinate with the appropriate person using those pandas to get them updated.

Any boards that fail the update will need to be reimaged with the preseed which will require hands on.

Comment 1

[Armen [:armenzg]]

Does this affect b2g pandas?
Racks 1 to 5 are allocated for b2g imaging (201-[1-5] - SCL1) and are not yet higly utilized.

Comment 2

[Jake Watkins [:dividehex]]

yes it does affect them

Comment 3

[Dustin J. Mitchell [:dustin] (he/him)]

Armen, remember that from our perspective there aren't "Android Pandas" and "B2G Pandas".  There are just "Pandas".

Comment 4

[Jake Watkins [:dividehex]]

I've update all pandas that were not in the locked_out state with the exception to the 198 that were allocated here (bug 826795)

Here is where the pool stands now

Pandas that failed:
0034 -  failed_pxe_booting
0106 -  pc_pinging
0250 - 	failed_pxe_booting
0258-290 - failed_pxe_booting
0342 - 	failed_pxe_booting
0468 - 	failed_pxe_booting
0472 - 	failed_pxe_booting
0488 - 	failed_pxe_booting
0554 - 	failed_pxe_booting
0562 - 	failed_pxe_booting
0570 - 	failed_pxe_booting
0574 - 	failed_pxe_booting
0598 - 	failed_pxe_booting
0621 - 	failed_pxe_booting
0662 - 	failed_pxe_booting
0720-0730 - pxe_power_cycling - possible relay failure -- see 821239
0741 -  failed_pxe_booting
0781 -  failed_pxe_booting
0793 -  failed_pxe_booting
0796 -  failed_pxe_booting

Pandas that are hardcoded and/or locked out:

0010-0021 - prototype
0022-0033 - releng staging + joel -- see also bug 816669 (LOCKOUT) hard coded
0046-0057 - kmoir (cedar tests in prod) (LOCKOUT) hard coded
0058-0069 - kmoir (cedar tests in prod) (LOCKOUT) hard coded
0808-0873 - hard coded - bug 826781

Comment 5

[Jake Watkins [:dividehex]]

These will be tracked in bug 817103 as failed pandas