From another node you can use isi_rdo

to execute commands remotely, it doesn't use ssh.

If the ssh server daemon is present but "just" hangs, try

isi_rdo "affectednodename" killall -v sshd

and wait for sshd to restart (maybe need killall -9 -v sshd).

If the sshd has gone and won't come back, call support.

hth

-- Peter

Hi Peter, thank you very much for the command. At the last time I used the command isi_for_array -nX killall -9 sshd. But it doesn't work. In the meantime, I opened a SR by EMC². If I get an anwer from EMC², I will post the result.

Thanks and have a nice day!

Hi Stefan,

Have you received the feedback from EMC support center? If not,  you can refer to the primus article "How to restrict SSH and OneFS web administration interface access to an Isilon cluster" to restrict the SSH connection to the issued node.

Hi Jeffey, thanks a lot for the link. Unfortunately, EMC² has not yet found a solution. At the Moment I'm still waiting for an feedback. I think, that two Services may cause the issue. SSHD and LSASS... I hope soon to hear from EMC². I will post the answer here :-)

Hi Stefan,

Thanks for your feedback. Looking forward to hear the case result from you.

Hi everybody, sorry for my late response.

Meanwhile I had several WebEx sessions with EMC². At the Moment the Problem is NOT solved .

Also I tried all of your commands.... But nothing has helped.

In a WebEx Session my active Directory account was added to the SystemAdmin role. After that the Problem was fixed. But this cant't be the solution.... It's just a workaround. Because I'm already member of a active Directory Group, that is member of the SystemAdmin role. In my opinion , I think that the affected node has a Problem to read the members of the active Directory Groups. There must be a Problem with a hanging service. Because a simple reboot will fix the Problem, without a manually adding to the SystemAdmin role.