Tim,

Don't be concerned that you're new to linux. We were all there at one point. ;)

Your input regarding your environment and what you've observed is helpful. I'll try to ask specific questions.

From what you've said, it appears that the problem exists in 2.6.9 (COS4/ESXi) kernels as well as 2.6.18 (COS5/Server2) kernels. There's fortunately some consistency in this.

I'm running all 32-bit, so the fact that you've seen it on 64-bit indicates that it's probably across both arch's. Is the QTP guest on this host also 64-bit, or is it 32-bit?

My present qmt-cos5 is also built from qmtiso5 and upgraded to 5.4, but 32-bit. I'd like to know, did you upgrade (all) other packages with qtp-newmodel and the unionfs sandbox? Your upgrade history should be evident from files in the /usr/src/qtp-upgrade/ tree if you don't recall. There are log files there as well as both source and binary rpms. Also, did you "yum update" the OS packages before or after you brought QTP up to date with qtp-newmodel?

You're the first person I recall who has run qtp-newmodel with a QTP guest on an ESXi host, although there are probably others who have done so. (Phil?) That could end up being very helpful.

As well as the host kernel versions, I think that the guest kernel version is perhaps more significant. What are the arch and kernel versions of the corresponding QMT-VMs that you have running?

Thanks.

--
-Eric 'shubes'

altoproxima wrote:
Eric,

The server version runs kernel 2.6.18.164.15.1.e15 on X86_64. It's the qmtiso5 upgraded to 5.4

the exsi 3.5 runs 2.6.9-89.0.23.ELsmp on i686




greetings

Tim

2010/4/18 Eric Shubert <[email protected] <mailto:[email protected]>>

    altoproxima wrote:

        Hello Eric,

        I am having exactly the same problem on both a Esxi 3.5 vmware
        and a Vmware server 2 setup on a ubuntu server 8.LTS

        Have you gotten around this problem allready ?


        greetings

        Tim


    Thanks for the note, Tim. This helps to confirm that the problem
    exists solely with VMware guests and the fuse-unionfs kernel modules.

    I'm seeing it with both COS4 and COS5 kernels. Also only with the
    latest clamav-toaster package thus far. I don't know if any other
    package do a similar chown command or not.

    Which kernel versions are you running on the hosts?

    It'd be nice to track this sucker down. I tried running the commands
    'manually' (chown into the sandbox, then run the pre script) and
    have not been able to cause the problem this way. FWIW, the command
    does appear to have done it's job. For some reason though the rpm
    pre script is detecting an unsuccessful exit code. Perhaps there's a
    non-zero return-code value being passed back from fuse-unionfs
    somehow when it's run in a guest?

    Do you have any ideas about how we can pinpoint the error?

-- -Eric 'shubes'


    
---------------------------------------------------------------------------------
    Qmailtoaster is sponsored by Vickers Consulting Group
    (www.vickersconsulting.com <http://www.vickersconsulting.com>)
      Vickers Consulting Group offers Qmailtoaster support and
    installations.
        If you need professional help with your setup, contact them today!
    
---------------------------------------------------------------------------------
       Please visit qmailtoaster.com <http://qmailtoaster.com> for the
    latest news, updates, and packages.
            To unsubscribe, e-mail:
    [email protected]
    <mailto:[email protected]>
       For additional commands, e-mail:
    [email protected]
    <mailto:[email protected]>





---------------------------------------------------------------------------------
Qmailtoaster is sponsored by Vickers Consulting Group 
(www.vickersconsulting.com)
   Vickers Consulting Group offers Qmailtoaster support and installations.
     If you need professional help with your setup, contact them today!
---------------------------------------------------------------------------------
    Please visit qmailtoaster.com for the latest news, updates, and packages.
To unsubscribe, e-mail: [email protected]
    For additional commands, e-mail: [email protected]


Reply via email to