494213 - opsi should be able ensure win32 machines have the same, consistent password

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P2)

Description

[[e-mail] (:bhearsum)]

Basically, we need a "password update" opsi package.

Comment 1

[[e-mail] (:bhearsum)]

Attachment: barebanes passwordupdate package

(I don't know how much review you can do on this, but another set of eyes never hurts.)

passwordupdate.ins will be in the opsi-binaries patch I'm about to post - we can't keep it somewhere public because it will have plaintext versions of our passwords in it.

Comment 2

[[e-mail] (:bhearsum)]

Attachment: passwordupdate.ins, passwords removed

Not much going on here - the DosInAnIcon and Registry parts of the function names are required by OPSI for what's happening in them. I tested this on win32-slave04 by changing the passwords a couple of times and it worked perfectly.

Comment 3

[Chris Cooper [:coop] (he/him)]

Comment on attachment 378894 [details] [diff] [review]
barebanes passwordupdate package

Is 0 the highest (i.e. most important) for priority? I'm thinking password updates are probably the most important thing we'd be doing on these boxes.

Comment 4

[[e-mail] (:bhearsum)]

(In reply to comment #3)
> (From update of attachment 378894 [details] [diff] [review])
> Is 0 the highest (i.e. most important) for priority? I'm thinking password
> updates are probably the most important thing we'd be doing on these boxes.

Good question. I'm pretty sure 'priority' just controls in which order things run, though. I'll look into it.

Comment 5

[[e-mail] (:bhearsum)]

(In reply to comment #4)
> (In reply to comment #3)
> > (From update of attachment 378894 [details] [diff] [review] [details])
> > Is 0 the highest (i.e. most important) for priority? I'm thinking password
> > updates are probably the most important thing we'd be doing on these boxes.
> 
> Good question. I'm pretty sure 'priority' just controls in which order things
> run, though. I'll look into it.

From the Manual: "[Product]-'Priority' is for future use (regarding the installation order)." - I think it's probably fine how it is for now.

Comment 6

[[e-mail] (:bhearsum)]

Comment on attachment 378895 [details] [diff] [review]
passwordupdate.ins, passwords removed

Checking in password-update/passwordupdate.ins;
/mofo/opsi-binaries/password-update/passwordupdate.ins,v  <--  passwordupdate.ins
initial revision: 1.1
done

Comment 7

[[e-mail] (:bhearsum)]

Comment on attachment 378894 [details] [diff] [review]
barebanes passwordupdate package

changeset:   5:642eabf15525

Comment 8

[[e-mail] (:bhearsum)]

This already got a good test run, so I'm going to declare it FIXED.

Comment 9

[Nick Thomas [:nthomas] (UTC+12)]

win32-slave04 did something a bit strange. nagios was reporting it was down for a few hours, and on connecting with VNC it wanted a login. Giving it Control-Alt-Delete (technically function-control-alt-backspace) resulted in it logging in without password prompt. It worked fine when I rebooted it again.

Comment 10

[[e-mail] (:bhearsum)]

(In reply to comment #9)
> win32-slave04 did something a bit strange. nagios was reporting it was down for
> a few hours, and on connecting with VNC it wanted a login. Giving it
> Control-Alt-Delete (technically function-control-alt-backspace) resulted in it
> logging in without password prompt. It worked fine when I rebooted it again.

That's weird. I *was* testing a Buildbot OPSI package there before I finished up with it. I wonder if it was fallout from that? Either way, I'll dig into this a bit and try to repro.