Hi, I’m Jesús (common name in Spain and South America, so no biggie here) and I usually live in London, although I’m originally from Spain. I say usually because now I’m taking a gap year to go travelling around the world.
If you have a look at my blog you’ll see that I’m writing about my travels, with photos I have on my flickr account: http://www.flickr.com/photos/golan. So feel free to wander around.
When not travelling, I write occasionally on this blog, about computers, life, stuff, anyway. Feel free to come by, leave a message, etc. I have another blog in Spanish, at http://blog.roncero.org.
J.
Nice Info 4 Tux Admins!!!
Keep more coming
Comment by koolhead17 — January 7, 2010 @ 17:14
Hello Jesus Roncero, I have been reading your blogs for sometime now. I am a fan of your writing. You know the way to explain complicated things in easy manner. I have been struggling with a situation for some time now. Please see if you can help or suggest anything.
I have VM running on OpenSUSE 11.4, Apache 2.2.17 and PHP 5.3.5. I am hosting moodle on it. Moodledata which is quite big directory almost 50GiB is on a NFS Server and Moodle Files is also on the same NFS Server. they are mounted on the VM and apache is installed on the VM which is serving these. Moodle also requires a DB which is MySQL running on different physical box.
After the deployment i am seeing some weird issue. CPU hiccups. Any CPU extensive work on the server bumps up the CPU like anything. This VM has 2 vCPUs. Still the issue is same. Even one log-rotation ran 5:00 PM EST and Load avg went to 200% usages. Apache was not able to serve the requests. then log-rotation reloaded apache still the website didn’t came up. Then the only option left was reboot the VM. Once we did that it all came to normal. In error_log i found warning saying child process did not exits, sending SIGNTERM. by default the Kernel was desktop type with OpenSUSE i updated it to default type. but still no improvement.
Can you think of something which can be causing this issue. few details:
NFS is running on clustered windows server.
monster#uname -a
Linux lms1056 2.6.37.6-0.5-default #1 SMP 2011-04-25 21:48:33 +0200 x86_64 x86_64 x86_64 GNU/Linux
VM has 10GiB RAM and 2 vCPUs
Comment by Satyajit — July 27, 2011 @ 17:33
Are there any issues between the VM and the NFS server, network-wise? perhaps it can’t read from it and then everything blocks trying to read from it???
Comment by jesus — August 2, 2011 @ 11:09
Our Architech did check the NFS server but no luck. But i have seen error like “lockd: server nfslms not responding, still trying” when the CPU is high. when i try nfsstat command it shows normal. if top also show that the MySQL server and nfs server are one where most of traffic moving aournd. problem is any CPU extensive work bumps up the load.simple logrotation ran on peak ours and bzip2 ran to zip the file for maximun 2 mins and CPU went of to AVg Load went 200, 175, 170. I have seen this much AVg Load first time in my life.
I am using
apache 2.2.17
PHP 5.3.5
Comment by Satyajit — August 3, 2011 @ 04:48