790240 - unable to log into talos-w8-hp-003-mgmt.winbuild.scl1.mozilla.com

Status: RESOLVED FIXED

(Infrastructure & Operations :: DCOps, task)

Description

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

The ilo on this machine is not accepting connections via ssh or http.  Could someone please investigate and see what's up?

Comment 1

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

bld-centos6-hp-034-mgmt.build.scl1.mozilla.com has the same issue.  Please reset the ilo on both of these so that the ssh and http interfaces both work.

Comment 2

[Derek Moore [:dmoore]]

arr, any chance we could get an ipmitools package installed on these? It would allow us to cold reset the iLO module without taking downtime for a server reboot.

Comment 3

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

These need to come down anyway, so it's not important for this particular bug (and they need to come down to patch the ilo).  For the future, though...

Is there a windows 8 ipmitool package?  I know that dustin had a bug open at one point to discuss adding hp utilities to these machines, but I don't think that ipmitool was added.

Comment 4

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

Since bld-centos6-hp-034.build.scl1.mozilla.com is getting reinstalled anyway, I went ahead and temporarily added the openipmi set to it and was able to do that one from the OS.  So just talos-w8-hp-003 needs attention now.

Comment 5

[Van Le [:van]]

Hi Amy,

I powercycled the host and I am able to access iLO now.

vle@vle-10516 ~ $ nc -nv 10.12.41.73 443
Connection to 10.12.41.73 443 port [tcp/*] succeeded!

Strange issue I noticed - using Firefox to access iLO via http, I am getting the following error with my laptop and Vinh's:

Your certificate contains the same serial number as another certificate issued by the certificate authority.  Please get a new certificate containing a unique serial number.

(Error code: sec_error_reused_issuer_and_serial)

However, using Safari, both our laptops can access iLO fine. 

Please let me know if you need further DCOPs assistance.

Thanks,
Van

Comment 6

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

That fixed the issue, thanks!

Comment 7

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

Van - that issue occurs with these HP's - I think because they have the same internal hostname, but I'm not sure.  The fix is to clear out your saved certificates, and when clicking through the security warnings, clear the checkbox to "save this exception".