Two identical machines, both running CentOS 4.3 plus all the errata.
First one:
yum update
went for a while and hung. At the bottom of the yum.log file was this:
Aug 31 06:58:01 Updated: control-center.i386 1:2.8.0-12.rhel4.5 Aug 31 06:58:09 Updated: iiimf-le-unit.i386 1:12.1-13.EL.5 Aug
The python process was in the process list, but it was consuming no CPU time. I let it go for a couple of hours, then killed it and played with the duplicate removal script posted previously.
Second one:
I tried a combination of suggestions previously posted:
yum clean all yum update yum sqlite python-sqlite python-urlgrabber python-elementtree yum clean all yum upgrade
This also ran for a while and then hung in a similar fashion. The tail of the yum.log is:
Aug 31 11:29:08 Updated: at.i386 3.1.8-80_EL4 Aug 31 11:29:09 Updated: samba-client.i386 3.0.10-1.4E.9 Aug 31 11:29:21 Updated: anaconda.i386 10.1.1.46-1.cento
Suggestions for future tries on other machines?
Thanks much, Zube
On Thu, Aug 31, 2006 at 12:32:35PM -0600, Zube wrote:
yum clean all yum update yum sqlite python-sqlite python-urlgrabber python-elementtree yum clean all yum upgrade
If the first part succeeded but then you get hangs in the second part, there goes my idea.
On Thu, 2006-08-31 at 12:32 -0600, Zube wrote:
Two identical machines, both running CentOS 4.3 plus all the errata.
<snip>
Suggestions for future tries on other machines?
My first attempt aborted, but at a much nicer place. Right at the end of the downloads. I rebooted and took the simplistic view (may apply, I was also 4.3 all updates and the occasional Extra or Plus goodie).
Feeling it was the typical Linux VM issue, I ran yum check-update and put the output into a list. I split it into two pieces (about 130 entries each, IIRC), made each an update script and ran one. It worked, pulling in dependencies as needed. I ran the other.
The only anomaly so far is iproute wanting to update again. I let it.
Thanks much, Zube
<snip sig stuff>
HTH -- Bill
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Thu, Aug 31, 2006 at 12:32:35PM -0600, Zube wrote:
Suggestions for future tries on other machines?
rm -f /var/lib/rpm/__*
and then start yum. It 'might' solve your problem.
[]s
- -- Rodrigo Barbosa "Quid quid Latine dictum sit, altum viditur" "Be excellent to each other ..." - Bill & Ted (Wyld Stallyns)
On Thu Aug 31 12:32:35 PM, Zube wrote:
Two identical machines, both running CentOS 4.3 plus all the errata.
First one:
yum update
went for a while and hung. At the bottom of the yum.log file was this:
Aug 31 06:58:01 Updated: control-center.i386 1:2.8.0-12.rhel4.5 Aug 31 06:58:09 Updated: iiimf-le-unit.i386 1:12.1-13.EL.5 Aug
The python process was in the process list, but it was consuming no CPU time. I let it go for a couple of hours, then killed it and played with the duplicate removal script posted previously.
Second one:
I tried a combination of suggestions previously posted:
yum clean all yum update yum sqlite python-sqlite python-urlgrabber python-elementtree yum clean all yum upgrade
This also ran for a while and then hung in a similar fashion. The tail of the yum.log is:
Aug 31 11:29:08 Updated: at.i386 3.1.8-80_EL4 Aug 31 11:29:09 Updated: samba-client.i386 3.0.10-1.4E.9 Aug 31 11:29:21 Updated: anaconda.i386 10.1.1.46-1.cento
Following up to my own message, upgrades to 4.4 have been much smoother with:
yum update yum sqlite python-sqlite ; yum update
I've yet to experience a hang using this. Many thanks to Karanbir Singh for the suggestion.
Cheers, Zube
On Fri Sep 01 02:19:45 PM, Zube wrote:
Following up to my own message, upgrades to 4.4 have been much smoother with:
yum update yum sqlite python-sqlite ; yum update
I've yet to experience a hang using this. Many thanks to Karanbir Singh for the suggestion.
I spoke too soon. This hung as well on another machine. A trace of the process reveals the dreaded futex:
futex(0xaa2cd28, FUTEX_WAIT, 1, NULL
The machines where this happened are all identical:
Gateway E-1800, Celeron 1300 MHz stepping 01, Maxtor 2B020H1 ATA, 512 MB memory
Zube
On Fri, Sep 01, 2006 at 03:28:47PM -0600, Zube wrote:
yum update yum sqlite python-sqlite ; yum update I've yet to experience a hang using this. Many thanks to Karanbir Singh for the suggestion.
I spoke too soon. This hung as well on another machine. A trace of the process reveals the dreaded futex: futex(0xaa2cd28, FUTEX_WAIT, 1, NULL
Okay, next theory:
1) upgrade the kernel first 2) reboot 3) upgrade python-sqlite 4) upgrade yum and sqlite and python-urlgrabber and whatever else 5) upgrade everything else