Disaster Recovery

xmats

Verified User
Joined
Dec 24, 2005
Messages
11
Hi everyone..

I'm in a world of pain right now and need some good advice regarding recovery of a server with DirectAdmin. I'm a DA newbie so I do not really have that much experience with DA yet.

This is the situation..

My server did not respond after a reboot. No services was available and I could not connect to SSH/web/anything. Before the reboot I changed the IP on the server and did an "yum update". After the reboot I was not able to connect to the server, but the server did respond to pings/traceroutes.

My datacenter (fully automated) has a control panel where I can boot a recovery OS into RAM and access the server. I did that, mounted the disks and deleted the firewall-rule which I initially thought must have been the problem. Another reboot, but still not able to connect to anything on the server (but still responding to pings/traceroutes). This leads me to believe that something is wrong with the Cent OS/Direct Admin install. If the OS was fubar, would the server even respond to pings? Why wouldn't services like SSH etc. start on reboot?

If I'm not able to figure this out, my only option left is a reinstall of CentOS/DirectAdmin and a restore of the accounts. I can access all files on the server via the recovery OS and I have a 100GB backup account at a backup provider which I can run the backup to. If I'm going down this route - how would I do this the most effective way? How can I pull all accounts (including home dir, sql and mail) and all DirectAdmin settings of the server and restore this later without being able to access any DirectAdmin tools?

Any help/insights is greatly appreciated. I'm also interested in paying for this kind of help from a DirectAdmin guru :)
 
When you changed the IP, did you change the IP on the DA license file as well? Did you setup the new routing information for the IP (if it is needed). Did you update all your apache httpd.conf for all the vhosts to list on the new IP instead of the old one? Did you change the IP your sshd listens on to the new one (if you have it listening on a specific IP)? Is the network files on startup initializing the new IP instead of the old one?
 
Marshall, I'd think that ssh would still be available, as long as the network was properly set up.

xmats, if you can get ssh running everything else should be easy to figure out and fix.

Do you have access to a KVM switch at the datacenter, or perhaps to an out-of-band connection to a terminal account through the bios (generally used for reloading software)?

My guess is there's a fundamental problem somewhere, and guessing where it is isn't going to be too practical. Even if we could fix it, we'd have to have access to fix it.

Feel free to contact me by email or telephone if you'd like (note that private messages are the last thing I read; please use email or telephone if you'd like to contact me).

Jeff
 
Back
Top