Discuss November 2005

discuss@lists.centos.org

232 participants
368 discussions

Squid updates break tunneling
by Tim Edwards 30 Nov '05

30 Nov '05

This is a question I posted to the RHEL3 mailing list but didn't get a response: "We are currently running squid-2.5.STABLE3-6.3E.9 on one of our RHEL3 proxy servers because updating to the latest (squid-2.5.STABLE3-6.3E.14) seems to break any kind of tunelling, ie. things like instant messaging ports and ssh using proxytunnel (http://proxytunnel.sourceforge.net/). Has anyone else experienced this problem with the Squid updates from RHN?" I was wondering if anyone running Centos has any advice/info on this? -- Tim Edwards

1 0

Yum / Up2date issues and mirror.centos.org
by Johnny Hughes 30 Nov '05

30 Nov '05

There have been some complaints recently concerning this, so I thought an explanation of the setup AND recommendations are in order. First, I want to say that we need more donated servers with fast (at least 100mbit) internet connections. If you are making lots of money using CentOS and are a hosting provider, please consider providing us a donated server so we can better distribute CentOS. Now on to the message :) ----------------------------------------- 1. mirror.centos.org is a rrdns of several machines, so that we can distribute the load for CentOS downloads across several servers. Distributed load is the only way we can serve more than 1 TiB per day for updates. (That's right, 1 TiB per day, ~30TiB per month, double the traffic of 3 months ago, 10x the traffic of a year ago). We had several servers totally maxed out serving updates during the 4.2 release cycle. Did I mention server donations :) There are other ways to distribute the load other than rrdns ... like a load balancer ... but these usually only work well with geographically co-located machines. Since all our machines for mirror.centos.org rrdns are donated and they are disbursed throughout the US and EU. They don't work well with the typical load balancing. 2. Since our machines are not geographically co-located, we need to come up with another method to use multiple servers to distribute load. We are currently developing an application that will use geoip (and public mirrors that we verify are updated: http://www.centos.org/modules/tinycontent/index.php?id=13 along with the donated mirror.centos.org servers) to generate a mirrorlist for each country identified by geoip. This mirrorlist should then have good, fast and geographically accurate servers for each person who does updates. We hope to have this system functioning properly before the release of CentOS-4.3 and CentOS-3.7. If you use the optional fastestmirror plugin, you will even get the fastest mirror in each repo from the mirrorlist. The fastestmirror plugin is currently under heavy development by the people at yum (in which CentOS is helping) ... so it is going to change in the next couple months, but in it's current form it works well for those not on proxy servers to pick the fastest mirrors from a list. http://lists.centos.org/pipermail/centos-devel/2005-November/000947.html 3. Another common problem is transparent proxy servers and yum. HTTP mirrors and yum don't work really well via transparent proxy servers, so if you are using one, your best bet is to pick an FTP mirror from our public external server list (see mirror link above) instead of mirror.centos.org. ------------------------------- So, what can you do now if you have problems ... or just want to help. First thing is to pick several close mirrors and not use mirror.centos.org in your /etc/yum.repos.d/CentOS-Base.conf file. There is an example of a repo with more than 1 mirror in the fastestmirror link above. If you have more than 1 mirror for each repo (I recommend that you do) ... and if you don't have to use yum via a proxy server that blocks outbound client connections ... then you should try the fastestmirror plugin above to pick your best mirror from the list at yum run time. We will continue to strive to deliver CentOS within our means as we continue to grow by leaps and bounds, but we need the help and support of our user community if we are going to be able to continue to provide the same stellar services that we have all grown accustomed to. That means people who use CentOS will need to support it ... that is what the Community ENTerprise Operating System (CentOS) is all about. Thanks, Johnny Hughes

10 50

RE: [CentOS] Building rpms
by Bowie Bailey 30 Nov '05

30 Nov '05

From: Craig White [mailto:craigwhite@azapple.com] > On Wed, 2005-11-30 at 12:53 -0600, Johnny Hughes wrote: > > > > What is the purpose of the LDAP upgrade ... if it is security you are > > worried about ... those get in there. > > > > See this: > > http://www.redhat.com/advice/speaks_backport.html > > > > When you start changing major components, you greatly reduce the > > stability of CentOS for yourself ... and you ruin the system > > interoperability. > ---- > I pretty much agree with that last statement - and could never > conceive of getting an rpm of openldap/servers/client from Fedora > and rebuilding it on RHEL/CentOS without it being really really > tough to build and not breaking anything. > > I think the general consensus on openldap message base is to build > everything in /usr/local from source, which in my case, I built db4 > (4.2.52+patches), kerberos, cyrus-sasl, openssl and then finally > openldap - all from source and it wasn't nearly as hard as I feared > and left RHEL stuff alone and didn't break anything. The information > that I used to do this all came from Quanah's web pages at > Stanford... > http://www.stanford.edu/services/directory/openldap/ I'm looking at doing that. I was just trying to stay with RPMs if possible so that I don't run in to dependency issues later when I try to install an RPM that requires openldap. > Perhaps a less painful method might be to use Buchan Milne's rpm's > which would do much the same and though they seem to be created for > Mandriva, apparently can build/install on RHEL (sorry, I don't have > a URL for this but you can either post to openldap list or search > their archives). Not a bad idea. Anyone tried this on CentOS4? > Lastly, perhaps the least painful method of all is the pretty much > turnkey packages available from symas... <http://www.symas.com> Interesting. I may look into this. > Now, generally Red Hat back port works well enough but if you are > going to make RHEL/CentOS the base of a large directory service... > 2.0.7-20 (CentOS3) and 2.2.13-4 (CentOS4) simply don't cut it for a > number of reasons. I stick with them on most of my installations > because the number of users and the extent of the demands that I put > upon ldap are pretty minimal but if you are going to have a > substantial investment in time/energy in ldap, > fahgettibouddit...install current. That's about what we determined. > Recognize that 2.2.30 (I believe) is still the latest categorized > 'stable' - 2.3.11 (and I think it is now up to 2.3.12) is discussed > and sometimes casually referred to as 'stable' - I don't think that > it has 'officially' been designated so. Actually, 2.3.11 is stable as of 10/18. Bowie

2 1

Re: [CentOS] why was LILO removed from centOS 4.2?
by Brian T. Brunner 30 Nov '05

30 Nov '05

The question remains unanswered: WHY was LILO removed? We/you/they removed *something that works*. What took its place does not work everywhere LILO does... a functionality loss. I hear folks saying "that loss is corrected in a future release" That's saying GRUB is *almost* ready to replace LILO, which has been removed prematurely. The question was "Hey, REDHAT, *WHY*?" Brian Brunner brian.t.brunner(a)gai-tronics.com (610)796-5838 >>> mail-lists(a)karan.org 11/30/05 10:55AM >>> Nielsen, Steve wrote: > I meant not a visible choice by default :). > > Thanks for the tip. I was not aware of making it visible in that way. > > So at the prompt you type: > > "boot: lilo" ? > boot: linux lilo - K _______________________________________________ CentOS mailing list CentOS(a)centos.org http://lists.centos.org/mailman/listinfo/centos ******************************************************************* This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept for the presence of computer viruses. www.hubbell.com - Hubbell Incorporated

4 3

RE: [CentOS] Building rpms
by Bowie Bailey 30 Nov '05

30 Nov '05

From: Johnny Hughes [mailto:mailing-lists@hughesjr.com] > On Wed, 2005-11-30 at 11:28 -0500, Matt Hyclak wrote: > > On Wed, Nov 30, 2005 at 11:16:08AM -0500, Bowie Bailey enlightened us: > > > From: Matt Hyclak [mailto:hyclak@math.ohiou.edu] > > > > > > > > On Wed, Nov 30, 2005 at 11:05:48AM -0500, Bowie Bailey enlightened us: > > > > > > > > > > I found an rpm from Fedora Core Development on RPMFIND. Will this one work > > > > > with Centos, or should I grab the source rpm and rebuild it? > > > > > > > > Rebuild from src.rpm. > > > > > > Ok, tried that... > > > > > > $ rpmbuild --rebuild openldap-2.3.11-3.src.rpm > > > Installing openldap-2.3.11-3.src.rpm > > > error: Failed build dependencies: > > > bind-libbind-devel is needed by openldap-2.3.11-3.x86_64 > > > libtool-ltdl-devel is needed by openldap-2.3.11-3.x86_64 > > > > > > Where do I find these two? Do they have different names with CentOS, > > > or do I have to get them from elsewhere? > > > > > > Thanks, > > > Bowie > > > > You'll probably have to get them from the FC Development repository as well. > > You're probably looking at the bind and libtool srpms, if I had to guess. > > If you're lucky, those two won't depend on too much else, otherwise you'll > > spiral quickly into dependency hell. > > > > > What is the purpose of the LDAP upgrade ... if it is security you are > worried about ... those get in there. > > See this: > http://www.redhat.com/advice/speaks_backport.html > > When you start changing major components, you greatly reduce the > stability of CentOS for yourself ... and you ruin the system > interoperability. The purpose is extra features. Mainly, we are having problems with reliability. Twice now, when the system has crashed, the database was corrupted. The scary thing is that both times the database was completely idle when this happened. Version 2.3 fixes a problem with the database checkpoint command and also allows for an automatic database recover on startup. Hopefully, this will at least allow us to recover more easily in the case of a crash. We are also looking at updating Berkely DB to see if that would help. Bowie

1 0

RE: [CentOS] Building rpms
by Bowie Bailey 30 Nov '05

30 Nov '05

From: Matt Hyclak [mailto:hyclak@math.ohiou.edu] > > On Wed, Nov 30, 2005 at 11:05:48AM -0500, Bowie Bailey enlightened us: > > > > I found an rpm from Fedora Core Development on RPMFIND. Will this one work > > with Centos, or should I grab the source rpm and rebuild it? > > Rebuild from src.rpm. Ok, tried that... $ rpmbuild --rebuild openldap-2.3.11-3.src.rpm Installing openldap-2.3.11-3.src.rpm error: Failed build dependencies: bind-libbind-devel is needed by openldap-2.3.11-3.x86_64 libtool-ltdl-devel is needed by openldap-2.3.11-3.x86_64 # yum install bind-libbind-devel libtool-ltdl-devel Setting up Install Process Setting up repositories update 100% |=========================| 951 B 00:00 base 100% |=========================| 1.1 kB 00:00 addons 100% |=========================| 951 B 00:00 extras 100% |=========================| 1.1 kB 00:00 Reading repository metadata in from local files Parsing package install arguments No Match for argument: bind-libbind-devel No Match for argument: libtool-ltdl-devel Nothing to do Where do I find these two? Do they have different names with CentOS, or do I have to get them from elsewhere? Thanks, Bowie

4 3

gdb segfault under ddd
by Michael Ubell 30 Nov '05

30 Nov '05

I am using CentOS 4 GNU gdb Red Hat Linux (6.3.0.0-1.63rh) GNU DDD 3.3.9 (x86_64-redhat-linux-gnu) I bring up ddd on a program compiled -g and type: list function_name gdb seg faults. If I bring up ddd on gdb and then run gdb with the same program an type list function_name everything works just fine.

1 0

Addition of Horde
by John Hinton 30 Nov '05

30 Nov '05

Johnny Hughes wrote: >CentOS Errata and Enhancement Advisory 2005:1130-1 > >CentOS 4 x86_64 Horde - Enhancement Update (Extras Only) > >We are pleased to add the Horde Application Framework as well as the >following horde projects to the CentOS extras repository: > >horde >imp-h3 >ingo-h3 >kronolith-h3 >mnemo-h3 >nag-h3 >gollem-h3 >turba-h3 > >Horde is an full featured webmail system for your server, please see >this link for details: > > > This is FANTASTIC news! Great selection of a great product. I've resisted to change to it simply due to the KISS mode of operation I try very hard to follow. A big pat-on-the-back to you folks! John Hinton

2 1

Re: [CentOS] why was LILO removed from centOS 4.2?
by Brian T. Brunner 30 Nov '05

30 Nov '05

>>> mail-lists(a)karan.org 11/30/05 11:57AM >>> Brian T. Brunner wrote: > The question remains unanswered: WHY was LILO removed? You seem to have missed the post where I highlighted the fact that it has not been removed. ***Yes, the "removed" was only virtual (making it invisible unless you have ***knowledge from elsewhere) not surgical. > > The question was "Hey, REDHAT, *WHY*?" you also seem to have the wrong forum for that question, please contact redhat at one of the following ***Anybody here know of a reason? Brian Brunner brian.t.brunner(a)gai-tronics.com (610)796-5838 ******************************************************************* This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept for the presence of computer viruses. www.hubbell.com - Hubbell Incorporated

2 1

X windows is taking 99 percent of cpu
by Jerry Geis 30 Nov '05

30 Nov '05

I am running centos 4.2 on a P4 3 GIG with 2 GIG ram. This machine has been fine for months. my laptop is fine. X is taking 99 percent of the cpu after some time. All I was doing was editing code. If I killall -9 X it comes back for a while then maxes out again. Any ideas on something to tweek??? THanks, jerry ------------ Tasks: 120 total, 2 running, 118 sleeping, 0 stopped, 0 zombie Cpu(s): 50.2% us, 0.0% sy, 0.0% ni, 49.8% id, 0.0% wa, 0.0% hi, 0.0% si Mem: 2074188k total, 742456k used, 1331732k free, 39704k buffers Swap: 19984840k total, 0k used, 19984840k free, 527876k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 5270 root 25 0 170m 21m 5808 R 99.9 1.0 7:26.25 X 1 root 16 0 2264 556 476 S 0.0 0.0 0:00.44 init 2 root RT 0 0 0 0 S 0.0 0.0 0:00.00 migration/0 3 root 34 19 0 0 0 S 0.0 0.0 0:00.00 ksoftirqd/0 4 root RT 0 0 0 0 S 0.0 0.0 0:00.02 migration/1 5 root 34 19 0 0 0 S 0.0 0.0 0:00.01 ksoftirqd/1 6 root 5 -10 0 0 0 S 0.0 0.0 0:00.10 events/0 7 root 5 -10 0 0 0 S 0.0 0.0 0:00.06 events/1 8 root 6 -10 0 0 0 S 0.0 0.0 0:00.00 khelper 9 root 15 -10 0 0 0 S 0.0 0.0 0:00.00 kacpid 33 root 5 -10 0 0 0 S 0.0 0.0 0:00.00 kblockd/0 34 root 5 -10 0 0 0 S 0.0 0.0 0:00.00 kblockd/1 44 root 20 0 0 0 0 S 0.0 0.0 0:00.00 pdflush 45 root 15 0 0 0 0 S 0.0 0.0 0:00.25 pdflush 35 root 15 0 0 0 0 S 0.0 0.0 0:00.00 khubd 47 root 13 -10 0 0 0 S 0.0 0.0 0:00.00 aio/0 48 root 9 -10 0 0 0 S 0.0 0.0 0:00.00 aio/1 46 root 25 0 0 0 0 S 0.0 0.0 0:00.00 kswapd0 122 root 25 0 0 0 0 S 0.0 0.0 0:00.00 kseriod 192 root 5 -10 0 0 0 S 0.0 0.0 0:00.00 ata/0 193 root 7 -10 0 0 0 S 0.0 0.0 0:00.00 ata/1 195 root 20 0 0 0 0 S 0.0 0.0 0:00.00 scsi_eh_0 196 root 19 0 0 0 0 S 0.0 0.0 0:00.00 scsi_eh_1 214 root 15 0 0 0 0 S 0.0 0.0 0:00.18 kjournald 1144 root 6 -10 1560 460 380 S 0.0 0.0 0:00.02 udevd 1348 root 18 0 0 0 0 S 0.0 0.0 0:00.00 shpchpd_event 1828 root 5 -10 0 0 0 S 0.0 0.0 0:00.00 kauditd 2029 root 5 -10 0 0 0 S 0.0 0.0 0:00.00 kmirrord 2030 root 5 -10 0 0 0 S 0.0 0.0 0:00.00 kmir_mon 2049 root 15 0 0 0 0 S 0.0 0.0 0:00.28 kjournald 2050 root 19 0 0 0 0 S 0.0 0.0 0:00.00 kjournald 2789 root 16 0 2408 596 508 S 0.0 0.0 0:00.06 syslogd 2793 root 15 0 3072 472 404 S 0.0 0.0 0:00.00 klogd 2804 root 16 0 2584 476 404 S 0.0 0.0 0:00.00 irqbalance 2815 rpc 16 0 2980 580 484 S 0.0 0.0 0:00.00 portmap 2835 rpcuser 16 0 2440 756 656 S 0.0 0.0 0:00.00 rpc.statd 2868 root 16 0 5048 1016 844 S 0.0 0.0 0:00.01 rpc.idmapd

3 2

← Newer
1
2
3
4
5
6
7
...
37
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Discuss November 2005