VMware Cloud Community
stevenfoo
Contributor
Contributor

ESXi hang and can't login using SSH once is a month

We had problem once in a month (random) with our ESXi 3.5. It is a installable ESXi 3.5 on our DELL PE 2950 will 10GB of RAM and 2 X quad core CPU.

We had about 3 of them.

We can't seem login though VI client or ssh using putty. Normal we don't have this issue.

All VM serves are still online running and user able to work on it. We tried to restart the VM network and agent, however it also failed.

We always need to reboot the host itself.

However this does not happen in our ESX 3.0 and 3.5 server.

The last message on the console was as below

.

0 Kudos
7 Replies
geddam
Expert
Expert

By any chance did you enable lockdown mode, while adding this host to Virtual Center, if yes, then uncheck the box, while adding it back to virtual center.

Above this..If lockdown mode is not enabled and you still have this issue..File a support request to find the bug details,If you have a valid contract with VMware.... This seems to be a known bug with this version.

Thanks,,

Ramesh. Geddam,

VCP 3&4, MCTS(Hyper-V).

Please award points, if helpful

Thanks,, Ramesh. Geddam,
0 Kudos
stevenfoo
Contributor
Contributor

We actually have the similar issue like the link below:

http://spininfo.homelinux.com/news/VMware_ESXi/2009/09/14/Keep_losing_remote_management._What_gives_

Our ESXi 3.5 is running on level U4

I read the similar but the solution seems to be for 3.5 U2.

http://communities.vmware.com/thread/155337;jsessionid=0A0AB320F6779017516AE50A68FE4CCB

Lockdown mode is disable by default on all host.

Any more ideas?

0 Kudos
DSTAVERT
Immortal
Immortal

I have run into this occasionally. Restarting the management agents does work, however, it isn't possible to get past this when you have "can not fork". At that point there aren't enough available resources to do much. My solution has been to shutdown the least important VMs from within the VM itself until I have enough resources and then restart the management agents. Once I have control back I migrate the VMs and restart the host.

-- David -- VMware Communities Moderator
0 Kudos
geddam
Expert
Expert

Can you generate logs and attach a transcript of hostd and vpxa logs, if you are not able to while host hangs...generate logs after you perform a reboot from VI Client....

Thanks,,

Ramesh. Geddam,

VCP 3&4, MCTS(Hyper-V).

Please award points, if helpful

Thanks,, Ramesh. Geddam,
0 Kudos
DSTAVERT
Immortal
Immortal

After a reboot the logs are lost. ESXi writes logs to a ramdisk. You can set up a syslog server and or have ESXi write logs to a datastore. You can do that from the VI client in configuration/software/advanced settings.

-- David -- VMware Communities Moderator
0 Kudos
stevenfoo
Contributor
Contributor

Are you able to login using SSH or the console back after you shutdown a VM itself ?

Which one do you login to restart the management agent.

Thanks

0 Kudos
marcelo_soares
Champion
Champion

The sfcbd are using all ram. Disable it after a reboot by doing:

/etc/init.d/sfcbd-watchdog stop

/etc/init.d/wsmand stop

/etc/init.d/slpd stop

Check your build number, the latest versions of ESXi 3.5 should not show this problem anymore. Also, you can disable CIM at the ESX advanced settings (have the same effect than the above steps).

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

Consider awarding points for "helpful" and/or "correct" answers.

Marcelo Soares