Downloading files issue with Linux guest OS on ESX4.0 U1

[rbos3]

Hey everyone,

I am almost pulling my hair out because of a problem I am having with CentOS 5.4 on a VMware ESX server.

When I try to download anything larger then 1mb from the internet it hangs

around 50% completion and the download speed collapses untill it's 0.

This happens when using wget, but also with yum(I don't know if yum uses wget to receive it's files tho).

Windows doesn't experience this problem on the same ESX server.

CentOS 5.4 should be supported on ESX 4.

I tried different vNics, re-checked and re-checked my network config, reinstalled CentOS, tried 32- and 64bit versions..

Sigh..

My switch/router doesn't show any packet errors or packet loss to the machine too.

When I was testing yesterday, I found out that downloading big files using FTP works like a charm.

To give you guys an idea of what the problem is 'showing' I will paste some output of my SSH session when trying to

download using yum or wget.

Download a 100mb.bin testfile which I can download without problems on a Windows machine.

<strike><strike>root@sdws03 tmp</strike></strike># wget www2.versatel.nl/100mb.bin<br/><br/>
--2010-03-16 10:49:31--  [http://www2.versatel.nl/100mb.bin]<br/><br/>
Resolving www2.versatel.nl... 62.58.50.219<br/><br/>
Connecting to www2.versatel.nl|62.58.50.219|:80... connected.<br/><br/>
HTTP request sent, awaiting response... 200 OK<br/><br/>
Length: 104857600 (100M) <strike><strike>application/octet-stream</strike></strike><br/><br/>
Saving to: `100mb.bin.2'<br/><br/>
<br/><br/>
1% [=&gt;] 1,474,144   --.-K/s  eta 12m 37s<br/><br/>

Now, when I try and download, for example, the Virtualmin repositories.

INFO - Installing dependencies using command: /usr/bin/yum -y -d 2 install bind bind-utils caching-nameserver httpd l-server mysql-devel postgresql postgresql-server rh-postgresql rh-postgresql-server logrotate webalizer php php-domdovecot spamassassin mod_dav_svn cyrus-sasl-gssapi mod_ssl ruby ruby-devel rubygems perl-XML-Simple perl-Crypt-SSLea<br /><br/>
...in progress, please wait...                                                                                NFO - /usr/bin/yum -y -d 2 install bind bind-utils caching-nameserver httpd postfix bind spamassassin procmail perl-ql-server rh-postgresql rh-postgresql-server logrotate webalizer php php-domxl php-gd php-imap php-mysql php-odbc phasl-gssapi mod_ssl ruby ruby-devel rubygems perl-XML-Simple perl-Crypt-SSLeay: Succeeded.<br/><br/>
INFO -  Succeeded.<br/><br/>
INFO - Installing Virtualmin and all related packages now using the command:<br/><br/>
INFO - yum clean all<br/><br/>
Loaded plugins: fastestmirror<br/><br/>
Cleaning up Everything<br/><br/>
Cleaning up list of fastest mirrors<br/><br/>
INFO - yum -y -d 2 install virtualmin-base<br/><br/>
...in progress, please wait...                                                                                [http://mirror.leaseweb.com/centos/5.4/os/i386/repodata/primary.sqlite.bz2:] [Errno 4] Socket Error: timed out<br/><br/>
Trying other mirror.                                                                                [http://software.virtualmin.com/gpl/universal/ust-virtual-server-theme-6.9-1.noarch.rpm:] [Errno 4] Socket Error: time<br/><br/>
Trying other mirror.                                                                                [http://software.virtualmin.com/gpl/universal/usermin-1.430-1.noarch.rpm:] [Errno 4] Socket Error: timed out<br/><br/>
Trying other mirror.                                                                                [http://software.virtualmin.com/gpl/rhel/5/i386/clamav-data-0.95.2-1.vm.el5.i386.rpm:] [Errno 4] Socket Error: timed o<br/><br/>
Trying other mirror.<br/><br/>

The last part of this output (Socket error timed out) appeared when I was running yum update the first time after I installed CentOS.

This is a clean install.

I tried setting the timeout and retry values of yum.conf to a higher value but this didn't solve the problem.

Now, I'm going to try and see if I can get a physical machine on the same switch as the

ESX server and see if the problem is at VMware, network or CentOS side.

My thoughts aren't going to the network side since all other machines on the same switch, host, router are working OK.

I have posted this same question on the CentOS forums since I don't know where the problem is caused.

Thanks!

René

[rbos3]

Anyone yet?

---

If you found this information useful, please consider awarding points for "Correct" or "Helpful". Thanks!

---