Server failing at random times - New Info
 

Hello, I have a problem with my Red Hat 9 server. I have setup the firewall to allow incoming traffic on ports: 80, 22, and 25. For webserver, ssh, and ftp. My server is not firewalled other than with its own rules, it is simply connected to the internet via a Cisco Catalyst 3500.

However, most of the time the webserver, ssh, and ftp work fine, but sometimes, I cannot reach anything. At these times it won't even accept a ping.

I am using a VA Linux server, and I was originally using the built-in LAN card, I thought that it might have been a source for the problem, so I replaced it with a known working LAN card. Yesterday, it was up for a good 8-12 hours with no problems, but no I cannot access it today.

It is really frustrating. I'm using Apache, sshd, and vsftp for my server. my ip for the server is 207.87.182.132 feel free to check it out and see what you come up with, any help would be appreciated.

Here are a few suggestions that I have received, but have not tried. I read somewhere that my network card might be setup to autonegotiate, and might be bringing the eth0 up and down because it is negotiating itself offline. ?? If I restart my server it will work fine, but restarting it 3 times a day isn't really what I'm looking for in a solution.

Also, I have access to the 3500 if there are certain configuration issues I might need to address, but a lot of the time, my server works fine.

This is my first post, so feel free to help out with any info. I will try to add more info as anyone needs it. Thanks.

-tobz

:newbie:

Have you been cleaning out old logs in /var/log? Is your /var partition low in space? If it starts to fill up, it will make the system very sluggish as the disk is hunting for the space it can use.

# How much free drive space
df -h

# Show disk usage by current directory and all subdirectories
du | less

# What takes up so much space on your box
# Run from the directory in question and the largest chunk shows up last
find $1 -type d | xargs du -sm | sort -g

# Redhat links
RedHat Linux Manuals
Get your mp3 support here
Maximum RPM
rpmfind
Easier software management: apt4rpm - Red Carpet
RedHat 8.0 Tips & Tricks

# Redhat 7.3 down configuration commands
setup leads to several configuration tools

# Redhat 7.3 up configuration commands
Configure soundcard:
redhat-config-soundcard
Configure X server:
redhat-config-xfree86
Configure network:
redhat-config-network
Manage software:
redhat-config-packages

# Handling NTFS
New Technology FileSystem (NTFS) HOWTOs
Linux NTFS project

Well, in /dev/sda2 (my / directory), there is 29G left, in /dev/sda1 (my boot dir), there is 15MB free out of 100 total, and in my swap partition, it says, filesystem none, size 1009M, used 0, avail 1009M, mounted on /dev/shm.

I believe that my drive space is probably not the issue, but you never know.

When I did the command to sort by largest dir, my largest dirs are /usr = 857M, and my halflife server /hlds = 742, and that accounts for most of my filespace.

I think it might be helpful to mention my hardware:
36GB 10k RPM SCSI Drive
Intel 880gx system board
2GB ram
dual PIII 850
floppy/cdrom

The server is UP now, not sure why, but I will probably be putting a monitor on it soon, so I can tell when it's down, and possibly link it to a reason. Thanks!

New Info - SSH ?
 

When I run my SSH client from my windows machines, sometimes it hangs when I don't exit right away. I also notice that after these hangs, my website runs slower and slower, and the SSH will not be available. The FTP usually will stop working after that, and the website will "eventually" go down.

I use putty.exe for my windows SSH client, should I use a different program? Or possibly just use telnet? Or if it does hang, is there anything I could do to fix it w/o restarting? Thanks in advance for any help, I've printed off all the Red Hat 9 manuals and found nothing of great relavence, searched using google but couldn't find a great match either. Again, thanks in advance.

tobz

Just thought that I would let everyone know that the problem is fixed. There is a different machine on my network that was eating up the Cisco 3500's resources, so when I fixed that, I matched the settings on the switch and my server to both FULL DUPLEX 100BaseTX, and things are working great.

Thanks to any and all who helped.

tobz