server goes day at 4am everyday

> my server stops responding at 4am everyday so I have to manually reboot it!

Cat You verify what applications are beeing run from crontab daily?
/etc/crontab
and content from /etc/cron.daily/

[(06:39 AM)][(root@dedicated)] [(~)] $ vi /etc/crontab
SHELL=/bin/bash
PATH=/sbin:/bin:/usr/sbin:/usr/bin
MAILTO=root
HOME=/

# run-parts
01 * * * * root run-parts /etc/cron.hourly
02 4 * * * root run-parts /etc/cron.daily
22 4 * * 0 root run-parts /etc/cron.weekly
42 4 1 * * root run-parts /etc/cron.monthly

[(06:40 AM)][(root@dedicated)] [(~)] $ cd /etc/cron.daily/

[(06:36 AM)][(root@dedicated)] [(/etc/cron.daily)] $ ll
total 164
lrwxrwxrwx  1 root root   28 Oct 13 21:03 00-logwatch -> ../log.d/scripts/logwatch.pl
-rwxr-xr-x  1 root root  418 Apr 26  2006 00-makewhatis.cron
-rwxr-xr-x  1 root root  276 Feb 21  2005 0anacron
yum.cron (END)
-rw-r--r--  1 root root  851 Sep 28 14:13 0fpusage
-rwxr-xr-x  1 root root   85 Oct 16 15:17 0virtualhosting.site1
-rwxr-xr-x  1 root root   86 Oct 16 15:32 0virtualhosting.site10
-rwxr-xr-x  1 root root   86 Oct 16 15:33 0virtualhosting.site11
-rwxr-xr-x  1 root root   86 Oct 16 15:34 0virtualhosting.site12
-rwxr-xr-x  1 root root   86 Oct 16 15:34 0virtualhosting.site13
-rwxr-xr-x  1 root root   86 Oct 16 14:19 0virtualhosting.site14
-rwxr-xr-x  1 root root   86 Oct 17 14:13 0virtualhosting.site15
-rwxr-xr-x  1 root root   85 Oct 16 15:20 0virtualhosting.site2
-rwxr-xr-x  1 root root   85 Oct 16 15:24 0virtualhosting.site3
-rwxr-xr-x  1 root root   85 Oct 16 15:25 0virtualhosting.site4
-rwxr-xr-x  1 root root   85 Oct 16 15:28 0virtualhosting.site5
-rwxr-xr-x  1 root root   85 Oct 16 15:30 0virtualhosting.site6
-rwxr-xr-x  1 root root   85 Oct 16 15:30 0virtualhosting.site7
-rwxr-xr-x  1 root root   85 Oct 16 15:31 0virtualhosting.site8
-rwxr-xr-x  1 root root   85 Oct 16 15:32 0virtualhosting.site9
-rwxr-xr-x  1 root root  241 Sep 28 14:12 1gen_analog_data
-rwxr-x---  1 root root  168 Sep 28 14:12 1hup_apache_logs
-rw-r--r--  1 root root  797 Feb 21  2005 certwatch
-rwx------  1 root root  257 Sep 28 14:15 check_quotas
-rwxr-xr-x  1 root root  998 Jun  1  2004 clean.quarantine
-rwxr-xr-x  1 root root  456 Jan  9  2006 freshclam
-rwxr-xr-x  1 root root  180 Aug 22  2005 logrotate
-rwxr-xr-x  1 root root  345 Jul 17 08:01 poprelay.custodiat.pl
-rwxr-xr-x  1 root root 2133 Dec  1  2004 prelink
-rwx------  1 root root   83 Sep 28 14:10 rm_oneclick
-rwxr-xr-x  1 root root  104 Aug 12 17:01 rpm
-rwxr-xr-x  1 root root  121 Aug 21  2005 slocate.cron
-rwxr-xr-x  1 root root  286 Feb 21  2005 tmpwatch
-rwx------  1 root root   68 Sep 28 14:10 vacuumdb
-rwxr-xr-x  1 root root   51 Sep 28 14:13 wltouch.daily
-rwxr-xr-x  1 root root  158 Aug 12 13:40 yum.cron

One of those scripts from /etc/cron.daily  is downing the server. Simply try running them one by one - maybe one of them will not return ;)

Hi,

This might as well be as a part of a hardware failure sch as RAM or Hard Disk failure. And since the system llocks yp I'll suggest you to reboot in memetest mode gram the grub and check your memory throughly. (IT seems to me that it is very likely to be a RAM error since the system locks up.)

Then you can fsck all your volumes.

how can I test it ?

- Just boot-off the install CD
- When theinitial prompt appears type "memtest86" enter
It will start the RAM test.

whats this script in /etc/cron.daily?

#!/bin/sh
/usr/sbin/chroot /home/virtual/site1/fst /usr/bin/run-parts /etc/cron.daily

It fires chained cron jobs from directory: /home/virtual/site1/fst/etc/cron.daily
Note - that they are run as root, therefore may cause damage as well.

this are the scripts in that dir!
I think one of the sites are taking down the server!

[(03:38 AM)][(root@dedicated)] [(/home/virtual/site1/fst/etc/cron.daily)] $ ls

00-makewhatis.cron  00updatestats  rmattachments

[(03:38 AM)][(root@dedicated)] [(/home/virtual/site1/fst/etc/cron.daily)] $ less 00-makewhatis.cron

#!/bin/bash

LOCKFILE=/var/lock/makewhatis.lock

# the lockfile is not meant to be perfect, it's just in case the
# two makewhatis cron scripts get run close to each other to keep
# them from stepping on each other's toes.  The worst that will
# happen is that they will temporarily corrupt the database...
[ -f $LOCKFILE ] && exit 0
trap "{ rm -f $LOCKFILE ; exit 255; }" EXIT
touch $LOCKFILE
makewhatis -u -w
exit 0

[(03:39 AM)][(root@dedicated)] [(/home/virtual/site1/fst/etc/cron.daily)] $ less 00updatestats

#!/bin/sh
# --------------------------------------------------------------------------
# Copyright (c) Ensim Corporation 2000, 2001
# ENSIM CORPORATION - ALL RIGHTS RESERVED

# --------------------------------------------------------------------------
# $Id: updatestats,v 1.13 2005/07/28 10:47:00 mborate Exp $
# $Name: epl-10-1-0-branch $
# --------------------------------------------------------------------------
APACHE_LOG=/var/log/httpd/access_log
/usr/lib/opcenter/analog/updatestats.pl /var/usage/web $APACHE_LOG >/dev/null 2>/dev/null
/usr/lib/opcenter/analog/updatestats.pl /var/usage/ftp /var/log/xferlog  >/dev/null 2>/dev/null

[(03:39 AM)][(root@dedicated)] [(/home/virtual/site1/fst/etc/cron.daily)] $ less rmattachments

#!/bin/sh
# --------------------------------------------------------------------------
# Copyright (c) Ensim Corporation 2000, 2001
# ENSIM CORPORATION - ALL RIGHTS RESERVED

# --------------------------------------------------------------------------

find /var/www/squirrelmail-attachments/ -type f -mtime +1 -exec rm -rf {} \;

> I think one of the sites are taking down the server!
Possibly - You have to scan all sites for their crons. The scripts You provided are nothing suspicious, hence running many such scripts concurrently will cause very high load average, and in fact I saw servers that under high LA stop responding. This can be the cause.

is there a way to run each script from daily cron every 2 mins after each other ?

Cron does run it's jobs sequencially anyway.
But the job itself may go into background - then problem arises. Maybe the sites crons are run in parallel?

run-parts is a shell script, so you could indeed modify it to kick off a job every 2 minutes.