These are essentially the same versions I have on my host.

This error is intermittent too. For some packages, the problem never occurs. The first error you posted is from the sandbox. This last run, everything installed ok in the sandbox, and the error didn't occur until the upgrade to the live system. It happens to different packages as well. Definitely something intermittent. Sometimes it happens, sometimes it doesn't. These types of bugs are terribly difficult to track down, but not necessarily impossible.

So either the file really *is* bad, or its not bad and the integrity check is falsely failing. Because we've observed a package installing successfully in the sandbox and failing later, my guess is the later (the check is falsely failing). The rpm file is probably good, but the rpm program erroneously thinks it's bad. Somehow.

I wonder, is there perhaps a special character unprintable character in the filename that we're not seeing? Kinda doubt it, so long as the stock settings are used.

I'm beginning to think this may be a hardware problem, like a hard drive or memory going bad.

What's the state of your HDD(s)? Do you have raid of any kind? Can you run smartctl on the drives? How much free space is there?

--
-Eric 'shubes'

On 10/24/2012 10:22 AM, G V wrote:
Here you go.
rpm -qa | grep rpm | sort
redhat-rpm-config-8.0.45-32.el5.centos
rpm-4.4.2.3-28.el5_8
rpm-build-4.4.2.3-28.el5_8
rpm-devel-4.4.2.3-28.el5_8
rpmforge-release-0.5.2-2.el5.rf
rpm-libs-4.4.2.3-28.el5_8
rpm-python-4.4.2.3-28.el5_8

Eric, thanks for your patience and continued help. This has become a
long journey.
George

On Wed, Oct 24, 2012 at 9:58 AM, Eric Shubert <[email protected]
<mailto:[email protected]>> wrote:

    Dang. They're identical, which they should be. That's good.

    So how does rpm manage to install the packages in the sandbox, but
    chokes on them (sometimes) when installing them in the base system?

    I don't think I've run this stuff on COS 5.8 yet. Has anyone updated
    these packages with qtp-newmodel on COS5.8? Please let us know if
    you have. I wouldn't expect a problem with that, but who knows?

    Let's have a closer look at your rpm packages:
    # rpm -qa | grep rpm | sort

    --
    -Eric 'shubes'


    On 10/23/2012 10:33 PM, G V wrote:


        On 10/23/2012 9:47 PM, Eric Shubert wrote:

            Ok, George. Let's keep digging. There's something weird
            going on here.

            # cd /usr/src/qtp-upgrade/RPMS
            # ls -l *.i686.rpm

        -rw-r--r-- 1 root root  307536 Oct 23 22:17
        courier-authlib-toaster-0.59.__2-1.3.10.i686.rpm
        -rw-r--r-- 1 root root  853017 Oct 23 22:28
        courier-imap-toaster-4.1.2-1.__3.10.i686.rpm
        -rw-r--r-- 1 root root 2402408 Oct 23 22:28
        qmailadmin-toaster-1.2.16-1.4.__0.i686.rpm
        -rw-r--r-- 1 root root   33686 Oct 23 22:10
        qmail-pop3d-toaster-1.03-1.3.__22.i686.rpm
        -rw-r--r-- 1 root root  664187 Oct 23 22:10
        qmail-toaster-1.03-1.3.22.__i686.rpm
        -rw-r--r-- 1 root root 1181851 Oct 23 22:09
        vpopmail-toaster-5.4.33-1.4.0.__i686.rpm

            # md5sum *.i686.rpm

        ae50283857793c87e937b69d88616e__e8
        courier-authlib-toaster-0.59.__2-1.3.10.i686.rpm
        da9ca074e8eb76206f4bcbf059b432__9d
          courier-imap-toaster-4.1.2-1.__3.10.i686.rpm
        439a9ded3cd44c4dbda4f91a9c0932__ed
          qmailadmin-toaster-1.2.16-1.4.__0.i686.rpm
        16c8295e982013ed8ab37a5ddf1d03__a8
          qmail-pop3d-toaster-1.03-1.3.__22.i686.rpm
        cb261cd2f41fd41658e55c560d5eb8__43
          qmail-toaster-1.03-1.3.22.__i686.rpm
        fffeb206bb811ed1c2cab7261b47b4__8b
          vpopmail-toaster-5.4.33-1.4.0.__i686.rpm


            Then, with sandbox still mounted:
            # cd /mnt/qtp-sandbox/usr/src/__redhat/RPMS/i686
            # ls -l *.i686.rpm

        -rw-r--r-- 1 root root  307536 Oct 23 22:17
        courier-authlib-toaster-0.59.__2-1.3.10.i686.rpm
        -rw-r--r-- 1 root root  853017 Oct 23 22:28
        courier-imap-toaster-4.1.2-1.__3.10.i686.rpm
        -rw-r--r-- 1 root root 2402408 Oct 23 22:28
        qmailadmin-toaster-1.2.16-1.4.__0.i686.rpm
        -rw-r--r-- 1 root root   33686 Oct 23 22:10
        qmail-pop3d-toaster-1.03-1.3.__22.i686.rpm
        -rw-r--r-- 1 root root  664187 Oct 23 22:10
        qmail-toaster-1.03-1.3.22.__i686.rpm
        -rw-r--r-- 1 root root 1181851 Oct 23 22:09
        vpopmail-toaster-5.4.33-1.4.0.__i686.rpm

            # md5sum *.i686.rpm

        ae50283857793c87e937b69d88616e__e8
        courier-authlib-toaster-0.59.__2-1.3.10.i686.rpm
        da9ca074e8eb76206f4bcbf059b432__9d
          courier-imap-toaster-4.1.2-1.__3.10.i686.rpm
        439a9ded3cd44c4dbda4f91a9c0932__ed
          qmailadmin-toaster-1.2.16-1.4.__0.i686.rpm
        16c8295e982013ed8ab37a5ddf1d03__a8
          qmail-pop3d-toaster-1.03-1.3.__22.i686.rpm
        cb261cd2f41fd41658e55c560d5eb8__43
          qmail-toaster-1.03-1.3.22.__i686.rpm
        fffeb206bb811ed1c2cab7261b47b4__8b
          vpopmail-toaster-5.4.33-1.4.0.__i686.rpm


            The file sizes and md5sums should be identical. I'm guessing
            they're
            not. Please post the results and we'll go from there.

            FYI, the commands
            # qtp-mount-sandbox
            # qtp-umount-sandbox
            can be run standalone on the cli in case you need one or the
            other.



        
------------------------------__------------------------------__---------
        To unsubscribe, e-mail:
        qmailtoaster-list-unsubscribe@__qmailtoaster.com
        <mailto:[email protected]>
        For additional commands, e-mail:
        qmailtoaster-list-help@__qmailtoaster.com
        <mailto:[email protected]>





    ------------------------------__------------------------------__---------
    To unsubscribe, e-mail:
    qmailtoaster-list-unsubscribe@__qmailtoaster.com
    <mailto:[email protected]>
    For additional commands, e-mail:
    qmailtoaster-list-help@__qmailtoaster.com
    <mailto:[email protected]>





---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to