Software/ Hardware failure caused chaos.
Our web site, written Perl/Mysql (Root Server running RedHat 7.3 - Apache/2.0.49 (Unix) mod_perl/1.99_13 Perl/v5.8.4 DAV/2 mod_ssl/2.0.49 OpenSSL/0.9.6b PHP/4.3.6 l) and hosted with some German company, went down for no reason. After 25 hours hosting provider managed to fix the problem. They told us that:
> Thank you for contacting us, regarding our phone conversation. I would
> like to mention here, that there was a slight problem with the RAM time of
> your server (hardware), and now its been set to the proper value. I have
> just tried pinging your server its working fine and I hope that this
> problem will not happen again in future.
Now then server started I have found out that:
1. system time was wrong (year 2003, November)
2. some scripts in /etc/init.d/ were missing (i.e. postfix, apache2 )
3. log files were missing
4. Web site works dead slow (top does show huge load) but we don’t have many customers browsing our site simultaneously.
Prior to downtime (1 month)I upgraded apache server to apache 2 with the mod_perl2 installed which gave us the substantial improvement (at lease 5 times – according to the apache server-status module) in speed and http request processing. Now after down time we find our server working much slower as compared to the previous version of apache (we had number of complaints from our regular customers).By analysing current configuration (dmesg) and checking hard drive (hdparm –t -T /dev/hda) I don’t see any reasons (sufficient memory resources, I disabled extra services, enough processor speed) why it is the case.
Please advise what could be the problem. I made a complain but provider swears that they did not change anything neither do I, but web site wors so slow so my boss is getting very upset with me
Thanks to all
Last edited by leshka_uk; 08-11-2004 at 06:15 AM.
|