485519 - write opsi package to deploy sshd on windows build machines

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P5)

Description

[Chris AtLee [:catlee]]

We should figure out a way to get sshd working on our windows build slaves.  It would make updating the slaves soooo much easier.

Comment 1

[alice nodelman [:alice] [:anode]]

I have filed pretty much the same bug for all the talos boxes. :)

Just like it, this goes in Future for now - though it would be awesome all around.

Comment 2

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

As we are rolling OPSI into production for build/unittest machines, do we still need this bug at all?

Comment 3

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

(In reply to comment #2)
> As we are rolling OPSI into production for build/unittest machines, do we still
> need this bug at all?

I would argue yes. Mundane tasks like cleaning up build dirs in advance of releases would be made much easier on Windows by not having to RDP in. This isn't something OPSI can handle AIUI.

Suffering through windows build dir cleanup on slaves for 3.5.4 right now. :(

Comment 4

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

(In reply to comment #3)
> (In reply to comment #2)
> > As we are rolling OPSI into production for build/unittest machines, do we still
> > need this bug at all?
> 
> I would argue yes. Mundane tasks like cleaning up build dirs in advance of
> releases would be made much easier on Windows by not having to RDP in. This
> isn't something OPSI can handle AIUI.

Correct, opsi/puppet and release automation do not handle this currently. 

I'd suggest that having this cleanup be part of release automation (bug#508403 is the real cure. If bug#508403 was fixed, would we still want to install sshd on the win32 slaves?



> Suffering through windows build dir cleanup on slaves for 3.5.4 right now. :(
Urgh - sorry.

Comment 5

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

Mass move of bugs from Release Engineering:Future -> Release Engineering. See
http://coop.deadsquid.com/2010/02/kiss-the-future-goodbye/ for more details.

Comment 6

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

Morphing summary. 

1) This is about writing an opsi package to deploy sshd onto windows 2003 build machines

2) This is for build machines only, explicitly not test machines.

Comment 7

[Syed Albiz [:salbiz]]

using kpyM sshd package with license key from bug 497487

Comment 8

[Syed Albiz [:salbiz]]

Attachment: [untested]add_sshd_opsi

Had to use autoit package to install via graphical installer. Tested autoit script, but need root staging-opsi powers to test properly through staging-opsi. In addition the license key from https://wiki.mozilla.org/ReferencePlatforms/Win64#SSH seems to be invalid when tested with the latest installer.

Comment 9

[Chris AtLee [:catlee]]

Comment on attachment 479436 [details] [diff] [review]
[untested]add_sshd_opsi

Need to figure out how to deploy the license key too.

Comment 10

[[e-mail] (:bhearsum)]

Comment on attachment 479436 [details] [diff] [review]
[untested]add_sshd_opsi

r+ assuming this tests fine in staging

Comment 11

[Syed Albiz [:salbiz]]

Attachment: [tested]patch_opsi_package

After thorough testing, I've made a few changes that get rid of the nag screen, and also added an auto-it based uninstall script.

Comment 12

[Syed Albiz [:salbiz]]

Attachment: [tested]patch_opsi_package

Added a section to remove the firewall exception on uninstall

Comment 13

[Rail Aliiev [:rail]]

Comment on attachment 480942 [details] [diff] [review]
[tested]patch_opsi_package

Looks good.

Comment 14

[[e-mail] (:bhearsum)]

Probably best to do this in a reconfig window. At the very least, needs-reconfig? will make sure this lands soon.

Comment 15

[[e-mail] (:bhearsum)]

Comment on attachment 480942 [details] [diff] [review]
[tested]patch_opsi_package

changeset:   66:49061b7f307d

Comment 16

[[e-mail] (:bhearsum)]

Landed the "binaries" too:
RCS file: /mofo/opsi-binaries/kpyM-sshd/kts.ini,v
done
Checking in kts.ini;
/mofo/opsi-binaries/kpyM-sshd/kts.ini,v  <--  kts.ini
initial revision: 1.1
done
RCS file: /mofo/opsi-binaries/kpyM-sshd/kts118.exe,v
done
Checking in kts118.exe;
/mofo/opsi-binaries/kpyM-sshd/kts118.exe,v  <--  kts118.exe
initial revision: 1.1
done

Comment 17

[[e-mail] (:bhearsum)]

Set to deploy on all win32 build machines:
mw32-ix-slave*
try-w32-slave*
w32-ix-slave*
win32-slave*

Comment 18

[John Ford [:jhford] CET/CEST Berlin Time]

(In reply to comment #17)
> Set to deploy on all win32 build machines:
> mw32-ix-slave*
> try-w32-slave*
> w32-ix-slave*
> win32-slave*

Awesome! Thanks!

Comment 19

[[e-mail] (:bhearsum)]

Haven't seen any fallout here, a few random tests confirm that this is deployed.