790007 - install nsclient++ 3.x.y on talos-w8-hp-{001..003}

Status: RESOLVED FIXED

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

Description

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

To help debugging the RDP crashes

Comment 1

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

* downloaded nsclient++-0.3.9-x64.msi from http://nsclient.org/nscp/downloads
* accepted defaults
* start screen, type "notepad" into nothing, shift-right-click notepad, click "run as administrator", enter password
* open c:\program files\nsclient++\nsc.ini
* paste nsc.ini from an existing w64 system, save

test:
"C:\Program Files\NSClient++\nsclient++.exe" /test 
   type "CheckCPU warn=80 crit=90 time=20m time=10s time=4"

start the daemon:
"C:\Program Files\NSClient++\nsclient++.exe" /start

from a nagios server,
nc -vz $hostname 5666

reboot and try the nc test again.

I did this on
 talos-w8-hp-001
 talos-w8-hp-002
I can't login to talos-w8-hp-003 either via -mgmt (conn refused on ports 80, 443, 22) or RDP (connects, long timeout, then auth fails).  That may need someone to remember what's up, or an onsite trip.

Comment 2

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

Sounds like this is the RDP lockup we're trying to solve.  That it also affects iLO suggests it's more than RDP!  This host will need hands-on, then.

Comment 3

[Amy Rich [:arr] [:arich]]

Hm, I hadn't heard that the RDP problem manifested as being unable to connect to the ilo, just the console via the ilo.  This does seem like a separate issue.  I'll file a bug with dcops to get the machine rebooted.

Comment 5

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

[[e-mail]1 ~]# nc -vz talos-w8-hp-003.winbuild.scl1.mozilla.com 5666
Connection to talos-w8-hp-003.winbuild.scl1.mozilla.com 5666 port [tcp/*] succeeded!
[[e-mail]1 ~]# nc -vz talos-w8-hp-002.winbuild.scl1.mozilla.com 5666
Connection to talos-w8-hp-002.winbuild.scl1.mozilla.com 5666 port [tcp/*] succeeded!
[[e-mail]1 ~]# nc -vz talos-w8-hp-001.winbuild.scl1.mozilla.com 5666
Connection to talos-w8-hp-001.winbuild.scl1.mozilla.com 5666 port [tcp/*] succeeded!

'course, I can't RDP to -001 or -002 right now, but hey, they're all up.  Nagios is green.