I've got a RH EL3 machine that has worked flawlessly for the past 8 months until recenty. I ran up2date the other day and it downloaded 30+ packages to install, and then stopped ( this is via the GUI btw ) and froze. The process ran for 15+ hours with no activity. I eventually rebooted the machine.

Upon doing so I found that several config files were replaced with defaults, such as named.conf and the iptables firewall rules. Now it appears that up2date will no longer run. If I run it via the GUI interface it downloads the packages successfully and then goes to the install phase where it will process a package and then stop. It sits on that display forever.

If I run it from the console, ie 'up2date php' it sits and does nothing.

[root@fe01www root]# up2date php

.. then if I do a 'ps aux | grep up2date' I see

root 32368 0.0 0.9 13300 7232 pts/2 S 19:02 0:00 /usr/bin/python -
u /usr/sbin/up2date php

Not sure whats going on here.... also, since that reboot several other odd things have happened. For example if I type iptables --list, it lists two or three rules.. pauses for 10 seconds, lists a few more, pauses.. etc. Normally it just dumps them out instantly.

Also, wget has gotten a little odd as well. For example if I type,

wget -r -l 1 -nd -A "*.dbi" <site here> it will resolve the hostname and begin to connect, pause for 30 seconds, and then begin working. I thought it was failing but it turns out its acutally working, just doing nothing for a period of time.

When I run 'top' I get:

19:14:15 up 3:44, 2 users, load average: 1.11, 1.37, 0.85
93 processes: 90 sleeping, 2 running, 1 zombie, 0 stopped
CPU states: cpu user nice system irq softirq iowait idle
total 5.1% 0.0% 1.5% 0.0% 0.0% 0.1% 93.0%
Mem: 768984k av, 642956k used, 126028k free, 0k shrd, 71860k buff
418604k actv, 149320k in_d, 8100k in_c
Swap: 1566328k av, 2172k used, 1564156k free 404232k cached

PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND
2871 root 15 0 10248 10M 6936 S 4.1 1.3 1:52 0 httpd
2375 mysql 15 0 11780 11M 2324 S 1.3 1.5 1:40 0 mysqld
3557 jason 15 0 1264 1264 912 R 0.1 0.1 0:00 0 top
1 root 15 0 516 516 456 S 0.0 0.0 0:03 0 init
2 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 keventd
3 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 kapmd
4 root 34 19 0 0 0 SWN 0.0 0.0 0:00 0 ksoftirqd/0
7 root 25 0 0 0 0 SW 0.0 0.0 0:00 0 bdflush
5 root 15 0 0 0 0 SW 0.0 0.0 0:21 0 kswapd
6 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 kscand
8 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 kupdated
9 root 25 0 0 0 0 SW 0.0 0.0 0:00 0 mdrecoveryd
16 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 ahc_dv_0
17 root 25 0 0 0 0 SW 0.0 0.0 0:00 0 scsi_eh_0
20 root 15 0 0 0 0 SW 0.0 0.0 0:07 0 kjournald
75 root 25 0 0 0 0 SW 0.0 0.0 0:00 0 khubd
1478 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 kjournald
2137 root 15 0 580 580 496 S 0.0 0.0 0:00 0 syslogd
2141 root 15 0 480 480 420 S 0.0 0.0 0:00 0 klogd
2167 rpc 19 0 580 580 504 S 0.0 0.0 0:00 0 portmap
2186 rpcuser 25 0 728 728 648 S 0.0 0.0 0:00 0 rpc.statd


Anyway, I'm out of ideas. Not sure whats causing these problems, but they all began happening after the failed up2date. Any comments, suggestions?

Thanks much!

Jason

Maybe your internet connection is getting a little lazy?

I've played around with Fedora a while ago and also noticed that up2date was very slow. Of course I read on their site (I made myself an account with them - free account) that for the first x days I get full bandwidth downloads via up2date for my computer, and after that I am stuck on some lesser speed with them, as I was not paying for those updates. But as you are on the paid-subscription Enterprise edition, and that it does download that probably is not the case. Just thought that might be of interest.

And for wget, you've tried a variety of sites, or mainly the redhat ones? Because that could tell you that the problem might be with your computer and not their servers. Try adding some flags to wget, such as -v (verbose). I don't know exactly what -S or --server-response (same things) do but the man says "Print the headers sent by HTTP servers and responses sent by FTP servers."

Hope this was useful :-/

crap - this thing got posted twice somehow

my bad

up2date has always been buggy for me. On both FC2, and FC3. I don't use it anymore. I just use yum instead. Just type yum update, and it'll update everything just as well as up2date, but without a pesky GUI, and without the pesky crashing.

If it did download some updates already you can try to upgrade them with rpm.

They will be in /var/spool/up2date.

rpm -Uvh /var/spool/up2date/*rpm

It's interesting, if I type iptables --list it will list the rules but list two, pause for 10 seconds or so, list some more, and keep going. Though if you look at the processes it isnt gobbling up any cpu power. What the hell is this thign doing? If it waits 10 seconds between listing iprules I wonder if its just waiting forever on up2date and timing out.

Ideas?

Also, per David's comment I went in to /var/spool/up2date and it has lots of rpm's inside there., but if I rpm -Uvh any of them it will sit there forever doing nothing. Example,

[root@geckobox up2date]# rpm -Uvh zip-2.3-16.1.i386.rpm

.. from another console if I ps aux | grep rpm I see:

root 16461 0.0 0.2 7480 1748 pts/1 S 21:23 0:00 rpm -Uvh zip-2.3-16.1.i386.rpm

It has been sitting for over 15 minutes now, its not doing anything. So I assume its more of a RPM issue than a up2date issue...

Any suggestions of where to go from here?

How about uninstalling it first? Maybe it's the -U argument...
Did that do anything different?

I think everything RPM related is fubar. Even if do something simple like: "rpm -qa | grep zip" it will sit there forever and do the same thing.

So whats causing RPM to stop, and lots of other system related shit to slow to a crawl? Ideas? I'm running out of them fast here. I have no idea wtf is going on!

Maybe you have two simultaneous instances of some rpm command running? I don't know if that is the case, because I haven't come across such a situation. But on my system there are rpmv and rpmq run during the night.. I have no idea what they do. Who knows, just give it a look-see I find this strange, too.

If you did a reset while updating, u may have corrupted the rpm database.

try
To see if an segmentation fault occurs, or if it freezes.
do
to rebuild your rpm database.

Well, the problem with iptables --list was that I forgot I had just received authorative control for DNS, so adding a -n to the iptables stopped it from resolving and fixed that. Once I had the reverse zone setup it worked like a charm.

As for the RPM database, I just ended up deleting all the files up2date downloaded, and wiping and rebuilding the rpm db. After a good hour of that it found lots of damaged shit and eventually got everything working. Now everything is working like a charm.

Thanks much everyone!

Hi, I am having the same problem. Just installed FC3, clicked on the flashing red icon in the system tray to update, and it happily downloaded LOTS of header files, then just apparently froze. No disk activity, no internet activity, little to no CPU usage. FC3 was running just fine, but the update was totally hung. I waited around for a long time, nothing happened, and I finally killed it.

Sounds like wiping and rebuilding the RPM database is a possible solution. I can see from this thread that I can rebuild via a "rpm --rebuilddb", but where is the RPM database stored? Should I wipe it out (delete all files) first, before doing the rebuild?

This is all pretty odd, and a disappointing intro to FC3. I have a completely new installation, which I haven't dicked around with at all (so it is just as the FC3 team intended) and something basic like a system upgrade is broken, "out of the box". Not good!

Thanks for any and all help - FC3 looks promising, and I would like to get this to work.

maybe not.
do :
man yum
to see the yum options.
I think something like
yum clean will clean the headers(but im not sure).

the rpm database may be corrupted if your machine frozes when installing the packages and not when downloading them.

Use yum instead of up2date which may be buggy.
The connection may stop if many people try to download from the same server.
This is why mirror sites exist. You should configure your system to download things from a fast mirror near you, and not use the official site. When people download 650MB of updates from the official site all together it's pretty obvious that there will be a problem.
Any way do
yum update
to update your system

for more info read the unofficiall fedora Faq

And something else...
if you delete all files of the rpm database, your system wont be able to know what is installed and what is not.

The header files you downloaded should be in /var/spoof/up2date if i remember well.
Do a search to find this out.
Also, no matter how fast your internet connection is the update will take many hours, since yum or up2date will have to find and fix all the dependencies in your system.