VMware

This Question is Answered

1 "correct" answer available (10 pts) 2 "helpful" answers available (6 pts)
1 2 Previous Next 25 Replies Last post: Nov 18, 2009 8:14 AM by BTtech  

vmware-hostd crashes repeatedly after upgrade to RHEL 5.4 posted: Sep 4, 2009 8:54 AM

Click to view dirkgf's profile Lurker 2 posts since
Sep 4, 2009

We're running a VMware Server 2.0.1 server on an RHEL 5 machine (x86_64, Dell T300 with a X3363 processor). Everything has been running fine until we've upgrade from RHEL 5.3 to RHEL 5.4. Now vmware-hostd crashes after some actions with SIGABRT. Reverting back to the old kernel does not seem to help. So I'd like to ask if anyone else is experiencing this issue.

I've opened up a Support Request at VMware, but since RHEL 5.4 is not yet supported I assume, the request will not be handled.

Click to view twwlogin's profile Novice 7 posts since
May 15, 2007
We saw the same thing. Was easier to just reinstall 5.1, the "supported" version of RHEL 5 for vmware 2.0.1. glibc was updated as well so maybe that had something to do with the SIGABRT messages. Dunno.
Click to view djzort's profile Lurker 2 posts since
Sep 15, 2009

Could this issue be related to vmware-hostd failing to find libc on a fork?

If i run the vmware-hostd command after it has crashed, from the command line like this...

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/lib/vmware/vmacore/

/usr/lib/vmware/bin/vmware-hostd -a -u /etc/vmware/hostd/config.xml

i cant crash it

however if i kill it and stop/start vmware. shortly there after vmware-hostd will crash again.


Click to view djzort's profile Lurker 2 posts since
Sep 15, 2009
on a box that works, the cpu is a core duo. on our xeon boxes vmware crashes.


the only difference in cpu extensions is the nx bit....

...

this proved to be incorrect. it crashed eventually.

Click to view mcssoftware's profile Novice 32 posts since
Dec 27, 2006
dirkgf wrote:
just for reference in case somebody else is having this issue. We seem to have been able to resolve the problem for us. Following the steps we performed.
  • Log on to your VMware host.
  • Create the directory /usr/lib/vmware/lib/libc.so.6
  • Log on to an RHEL 5.3. machine. Grab /lib64/libc-2.5.so (or
    /lib/libc-2.5.so in case you're running an 32 Bit host) and copy it to
    the VMware host into /usr/lib/vmware/lib/libc.so.6
  • Rename the file libc-2.5.so within /usr/lib/vmware/lib/libc.so.6 to libc.so.6
  • Open /usr/sbin/vmware-hostd and add /usr/lib/vmware/lib/libc.so.6 to the LD_LIBRARY_PATH. I just added an "export LD_LIBRARY_PATH=/usr/lib/vmware/lib/libc.so.6:$LD_LIBRARY_PATH" before the last line.
  • Restart your vmware services (our just the host)

Downgrading libc might be also an option, but I'd like to keep the host unchanged otherwise.


I just wanted to post to this thread to bump this (wish there was a WIKI for this kind of thing). My situation was a bit different since I updated glibc on my box to add a dev environment to compile a server monitoring agent - little did I know (or realize) that this hosed vmware hostd until I rebooted the server a month later :(

I really wish vmware would just include their own versions of glibc and put that in path instead of relying on the OS. Library hell (DLL hell whatever) ain't fun. OR better yet put out a patch when RHEL is reved since it is a supported platform. Then again, I guess this is their way of saying a native hypervisor like ESX/ESXi is the better option to avoid these issues.


Anyway, It was great that you posted this procedure after discovering the issue. It definitely saved me (and hopefully others) a lot of time and stress. Thanks.

Click to view seifried's profile Novice 6 posts since
Oct 29, 2009

Just some additional info: to get a copy of glibc-2.5 from Red Hat/CentOS 5.3:

  1. rpm -qf /lib64/libc-2.5.so
glibc-2.5-42

Either download the RPM for glibc or copy it off your DVD/iso image/etc:

mount -o loop /mirrors/CentOS/5.3/CentOS-5.3-x86_64-bin-DVD.iso /mnt/iso/

Install glibc package into /tmp so we can get the file quickly (--root=/tmp/)

cd /tmp/
rpm -Uvh --root=/tmp/ --nodeps /mnt/iso/CentOS/glibc-2.5-34.x86_64.rpm

And you will now have the file in:

/tmp/lib64/libc-2.5.so

You can also use tools like* *rpm2cpio to conver the RPM package to cpio and then extract it from the cpio but installing it to /tmp is very quick and easy.


Click to view seifried's profile Novice 6 posts since
Oct 29, 2009

So after trying this it worked for a minute or two before crashing. Personally I think th best thing to do for now is downgrade to CentOS 5.3 and add the line:

exclude=glibc glibc-common glibc-devel glibc-headers nscd

to your yum.conf and wait this thing out. Ugh. Any eta on a fix?

Click to view axon-obriend's profile Lurker 2 posts since
Nov 20, 2003

OK... I just got bit in the a-- by this one and wanted to express my heartfelt thanks to the guy who posted the fix.


Click to view seifried's profile Novice 6 posts since
Oct 29, 2009

You're welcome, I love the fact that a few sentences can save people hours of work/stress.

Is there any chance of an official/semi-official statement from VMware on when this might be fixed?

-Kurt

Click to view knnniggett's profile Lurker 2 posts since
Nov 5, 2009
dirkgf wrote:
just for reference in case somebody else is having this issue. We seem to have been able to resolve the problem for us. Following the steps we performed.
  • Log on to your VMware host.
  • Create the directory /usr/lib/vmware/lib/libc.so.6
  • Log on to an RHEL 5.3. machine. Grab /lib64/libc-2.5.so (or
    /lib/libc-2.5.so in case you're running an 32 Bit host) and copy it to
    the VMware host into /usr/lib/vmware/lib/libc.so.6
  • Rename the file libc-2.5.so within /usr/lib/vmware/lib/libc.so.6 to libc.so.6
  • Open /usr/sbin/vmware-hostd and add /usr/lib/vmware/lib/libc.so.6 to the LD_LIBRARY_PATH. I just added an "export LD_LIBRARY_PATH=/usr/lib/vmware/lib/libc.so.6:$LD_LIBRARY_PATH" before the last line.
  • Restart your vmware services (our just the host)

Downgrading libc might be also an option, but I'd like to keep the host unchanged otherwise.

I just ran into this problem on my CentOS 5.4 server. After a bit of Googling I came upon this thread. Since applying your suggestion, the problem has not reoccured.

I'm glad I didn't have to downgrade anything to get this fixed.

Thank you for posting this solution!

Click to view Slapper153's profile Lurker 3 posts since
Jul 21, 2009
Hi !!

I would like to ask if someone check the same solution for vmware server2.0.2 and centos 5.4..

Thanks In advance !!


edit : i am currently running cento5.3 64bit with vmware server 2.0.1 with 4 vms..

Click to view Edward78's profile Lurker 1 posts since
Nov 6, 2009
Good time of day ...

The problem is not in the libraries. The problem in the configuration. I in particular, the problem was in the file / etc / vmware / hostd / authorization.xml, setting <NextAceIdn> n, while ACEData id = n already. In general, simply write the value n +1 and restart service.

Sorry it's libraries 8-/

Link: http://blog.olegk.ru/2009/10/centos-54-vmware-server.html

Translate:
Updated here on virtually all servers with Centos 5.3 to 5.4. Ogreb problem on two servers, where swirling Dev on vmware server 2.0.1 - Dev work, and the Web interface vmware does not work. It turns out this is a known bug, the problem in the interaction of vmware and the new version glibc.
Decision of the comments to the Bagua did not suit me - did not work for some reason. In shoved the old library vmware-hostd went happily to fall.
I had to roll back to glibc. And just yum downgrade glibc refused to roll back, saying that Nizzam. But when they download the old rpm-s (http://mirror.centos.org/centos/5.3/updates/x86_64/RPMS/glibc-2.5-34.el5_3.1.x86_64.rpm, http://mirror.centos .org/centos/5.3/updates/x86_64/RPMS/glibc-common-2.5-34.el5_3.1.x86_64.rpm, http://mirror.centos.org/centos/5.3/updates/x86_64/RPMS/glibc -devel-2.5-34.el5_3.1.x86_64.rpm, http://mirror.centos.org/centos/5.3/updates/x86_64/RPMS/glibc-headers-2.5-34.el5_3.1.x86_64.rpm ) and set them through all the same "yum downgrade
glibc-2.5-34.el5_3.1.x86_64.rpm glibc-common-2.5-34.el5_3.1.x86_64.rpm glibc-devel-2.5-34.el5_3.1.x86_64.rpm glibc-headers-2.5-34 . el5_3.1.x86_64.rpm ", yes reran vmware - all earned.
Well, and then added to the exceptions glibc yum, that has not been updated - in / etc / yum.conf added "exclude = glibc, glibc-common, glibc-devel, glibc-headers".
Click to view knnniggett's profile Lurker 2 posts since
Nov 5, 2009
Slapper153 wrote:Hi !!

I would like to ask if someone check the same solution for vmware server2.0.2 and centos 5.4..

Thanks In advance !!

edit : i am currently running cento5.3 64bit with vmware server 2.0.1 with 4 vms..

Just to clarify, I am running 64bit CentOS 5.4 with vmware server 2.02 and as previously posted the solution worked for my system.


Update: I checked my virtual machine this morning and it is still running since applying the "fix". I am confident this issue has been resolved on my system.

Click to view Slapper153's profile Lurker 3 posts since
Jul 21, 2009

ok !! thanks for the reply !!

Probably tomorrow i will give a try ! :p

VMware Developer

SDKs, APIs, Videos, Learn and much more in the Developer community.

Learn More

Developer Sample Code

Increase your developer productivity with VMware API sample code.

Learn More

VMworld Sessions & Labs

Online access to the latest VMworld Sessions & Labs and online services.

Learn more

Purchase PSO Credits Online

Purchase credits to redeem training and consulting services online.

Buy Now

Community Hardware Software

View reported configurations or report your own.

Learn More

VMware vSphere

Come witness the next giant leap in virtualization.

Register Today

Communities