s00pcan
Contributor
Contributor

"The server took too long to respond" when connecting to the vSphere client

I've had this error before, it probably means that the guest vms are using up too much memory/processing power. But I can't get in to change any settings or shut individual ones down. How can I do this from the service console? Or are there any other options?

0 Kudos
14 Replies

You probably are running low on memory on your ESX. Try restarting management agents from the alt+f2 screen, physically on the box. Try logging in again.

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

Marcelo Soares
0 Kudos
s00pcan
Contributor
Contributor

Actually, I had already tried that. But I restarted, tried it again. Still giving the same error.

0 Kudos

If you don't have a chance to reboot the host (I suspect not Smiley Happy ) try this (ESXi 4.0):

- Enter on it in unsupported mode (alt+f1 on the physical screen, type "unsupported" and type the root password).

- If you are able to enter, type the following commands:

/etc/init.d/sfcbd-watchdog stop

/etc/init.d/wsmand stop

/etc/init.d/slpd stop

Wait a minute and try to connect to the server.

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

Marcelo Soares
s00pcan
Contributor
Contributor

/etc/init.d/sfcbd-watchdog stop

/etc/init.d/wsman stop

/etc/init.d/slpd stop

/etc/init.d/wsmand didn't exist, had to use wsman instead. Why is there no easy way of starting/stopping vms from the host machine? Is that just esxi or does esx have that ability? Anyway, I just tried it again and it worked. Thanks.

0 Kudos

After the commands, try killing the hostd process. If you do a "ps aux | grep hostd" you will see a lot processes, and the second column is the parent PID. Do a "kill -9 PID" until the process is dead - probably it will start again automatically because of the watchdog. Repeat the kill process for sfcbd processes.

You can kill vms, but this is like a hard poweroff, and may harm your system. To be able to shutdown VMs, unfortunately hostd process should be working ok, does not matter if you are on ESX classic or ESXi.

If you want to kill directly your VMs, do a "ps aux | grep VMNAME", where VMNAME must be the name, case sensitive, of your vmx file. Take the PID of the process returned and kill it (usually 3 processes).

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

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

Marcelo Soares
s00pcan
Contributor
Contributor

New problem: all of my VMs are showing as Unknown (inaccessible) on the vSphere client now. It lists the 4 vms I had, but says there is no data store. I've had this happen before as well, I go to re-add the data store and it never finishes loading on the "current disk layout" step.

0 Kudos

Seems storage connection issue. Are you able to see your datastores?

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

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

Marcelo Soares
0 Kudos
s00pcan
Contributor
Contributor

No, trying to add/view/do anything with a data store just makes it hang at a loading screen forever. This has happened to me twice before, and both times I just gave up and reinstalled esxi and reinstalled all my vms. I'm trying to get an understanding of how to deal with issues like these ahead of time so when I start using this in a production environment I don't have to worry, so I really don't want to go through that again.

0 Kudos

You are using SAN or local disks? Check at your /var/log/vmkernel file to check for any errors you may be experiencing. Also, call hardware support for a health check.

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

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

Marcelo Soares
0 Kudos
s00pcan
Contributor
Contributor

It's just a local sata hard drive. I'm seeing now that this "add storage" wizard just takes forever to do anything, but does work. It's telling me my hard drive is blank and that I need to format it... why? What happened here that made the hard drive is suddenly unrecognizable by esxi? I didn't lose any important data if it is unrecoverable, but you can imagine why I'm learning this ahead of time now. /var/log/vmkernel doesn't exist; /var/log # ls

configRP.log messages sysboot.log

ipmi smbios.bin vmware

messages: http://pastebin.com/m3f96fece

sysboot.log: http://pastebin.com/m6d55e901

0 Kudos

What hardware you have? maybe it is not listed on the VMware HCL....

I told you the wrong filename, for ESXi you need to check /var/log/messages

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

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

Marcelo Soares
0 Kudos
s00pcan
Contributor
Contributor

It's a Seagate st3250310AS 250GB sata hard drive. Motherboard is a little over a year old, intel. I've been using it for weeks now in this system for esxi though.

0 Kudos

That's bad, if the hw is not on HCL it's hard to point the finger to the VMware product. Check the file I told you. But I am certain that, if you run the ESX on a certified hardware, problems like this will not happen, sure unless the hardware have some problem.

Marcelo Soares

VMWare Certified Professional 310/410

Technical Support Engineer

Globant Argentina

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

Marcelo Soares
0 Kudos
s00pcan
Contributor
Contributor

If it doesn't support a basic sata setup then.. wow. I get that this product is intended for high end servers, but come on. I guess I won't be using esxi then.

0 Kudos