Help: Server gone down.

[greg]

Hi, last night our server failed and I'm fast running out of ideas on how to solve it. I've been going through the error log and I've come across the following problems, I was hoping someone could guide me on how to deal with them:

1) [warn] NameVirtualHost 216.127.80.23:80 has no VirtualHosts

(there's 1000s of that one)

2) Failed loading /usr/local/Zend/lib/ZendOptimizer.so: /usr/local/Zend/lib/ZendOptimizer.so: cannot open shared object file: No such fil$

(to do with PHP, so I dont really believe this is doing to much harm.. however the machine is rebooting after this error so maybe its indirectly causing problems.)

3) [Tue Aug 13 22:33:42 2002] [error] [client 66.17.17.240] File does not exist: /usr/local/apache/htdocs/scripts/..%5c%5c../winnt/system3$

[Tue Aug 13 22:33:53 2002] [error] [client 66.17.17.240] File does not exist: /usr/local/apache/htdocs/scripts/..%5c%5c../winnt/system3$

[Tue Aug 13 22:33:53 2002] [error] [client 66.17.17.240] File does not exist: /usr/local/apache/htdocs/scripts/..%5c%5c../winnt/system3$

[Tue Aug 13 22:33:53 2002] [error] [client 66.17.17.240] File does not exist: /usr/local/apache/htdocs/scripts/..%5c%5c../winnt/system3$

[Tue Aug 13 22:33:53 2002] [error] [client 66.17.17.240] File does not exist: /usr/local/apache/htdocs/scripts/..%5c%5c../winnt/system3$

[Tue Aug 13 22:39:23 2002] [notice] child pid 2397 exit signal Segmentation fault (11)

[Tue Aug 13 22:42:02 2002] [notice] SIGUSR1 received. Doing graceful restart

4) [Tue Aug 13 21:01:36 2002] [crit] (98) Address already in use: make_sock: could not bind to port 443

5) [Tue Aug 13 20:19:07 2002] [error] [client 195.92.67.67] script not found or unable to stat: /usr/local/apache/cgi-bin/dasp

I'm really sorry to put all these errors in front of you all, but I dont know which one could be causing the problems.

Everything works when you access things via the IP address but when using domain names it fails. Another thing worth noting is not all domains when at the same time.. most went at 1 time, then the remaining few functioning domains eventually failed.

I realise everyone out there doesn't owe me a thing, but I'd greatly appreciate everyone's prompt help.. I'm sure you all know what its like with server down time.

Thanks to everyone in advance.

Greg

 

[kwimberl]

Fix zend problem with this:

/scripts/installzendopt

Then:

/etc/rc.d/init.d/httpd stop
/etc/rc.d/init.d/httpd startssl

 

[ecoutez]

Hrmm - similar problems last night...

Hey Greg,
I had some pretty similar problems last night that I, perhaps incorrectly, attributed to my efforts to get an SSL cert installed. (I was doing it manually as I have on other mod_ssl boxes, and CPanel went and removed my configuration settings from httpd.conf a little before midnight EST. A little scary - and here I though *I* was the boss!).

After CPanel did its work on the httpd.conf file, I got the same &no VirtualHosts& error while starting Apache. In my case, the NameVirtualHost that CPanel complained about wasn't even present. I found that the errors reported to WHM while attempting to start Apache were misleading at best. Trying a &/usr/local/apache/bin/apachectl configtest& as root gave me a completely different error... a missing &/VirtualHost& if I remember correctly. And that was a quick and easy fix.

I'm also now failing to load the ZendOptimizer - same exact error that you have.

No stability problems though - box is up and running.

- Jason

 

[greg]

kwimberl, thanks for the info! It hasn't solved our biggest problem but I'm confident its got our server closer to running like normal.

ecoutez, is there any chance you could tell us what you did to fix the problem? If you want you can email me at [email protected] - although I'm getting to grips with Linux please take into account I'm still a little rough around the edges..

Thanks in advance!

 

[ecoutez]

I'm pretty sure that the configtest showed that I had a missing &/virtualhost& directive and indicated the problem line... so then it was just a matter of editing the config file /usr/local/apache/conf/httpd.conf to correct the problem.

Greg - just sent you mail with a little more detail.

- Jason

 

[kwimberl]

4) [Tue Aug 13 21:01:36 2002] [crit] (98) Address already in use: make_sock: could not bind to port 443


Focus on that error. That's the one giving you problems at this point I think.

Port 443 is the SSL port.

If you can't figure it out, call or write your distributor. If they won't help you, you can find me on AIM (kwimberl) and I'll see what I can do.

 

[Annette]

Easy enough to take care of if apache won't restart and gives that error.

netstat -anp | grep 443

Look for whatever process is still binding to that port and kill it. You may need to do this a couple of times, since other processes will tend to latch on to the port as well. Once the above returns nothing, restart apache.

 

[ecoutez]

DNS server(s) down

Greg - I responded to your email. Check your authoritative DNS server(s) for both your primary domain and your clients. Neither IP address is responding to DNS queries (can't ping, can't traceroute, and no named daemon listening). This server isn't currently configured for those two IP addresses... so either it needs to be, or you need to check out the server that is supposed to be serving up the DNS zone files.

- Jason