Hi guys

So here is a "what would you do" question. It is based on reality , unfortunately :(

A servers hard drive crashes... it is replaced and a back up from a few days before the crash re-installed. More than 24 hours later the server loads are still sometimes reaching dangerous levels, email is erratic and there are some crashes

It is not exim or any of the obvious stuff to look at....and still clueless as to cause

To make matters worse it is a new server and mostly resellers who just moved their from a different server.

How would you handle this sort of thing?

Best

I'd reccommend upgrading to a more powerful server, and if you already arent there, go scsi. It can handle more data in/out. Also, get daily backups and if possible, a mirror server. There are some hosts that can have 100% uptime because of a mirror server ready to kick in if the primary one goes out.

You mean Raid1?

Server is SCSI with IDE backup it is non-RAID

Is it a leased server?

I'd get another one and move everyone again and leave that broken server far behind as quickly as possible.

No time to dilly dally in these situations. Start moving accounts and getting dns together. Otherwise, you're going to be babysitting some broken piece of junk day and night. You can move people quickly and easily if you've got root on both machines.

I am not positive if it is owned or leased by the host (Chris and I are resellers)
The host seems committed to fix this one.

I am nervous that even if they get it "stable" it may not really be stable...

I ran a linux redhat box and though I know little ..VERY little still ..I know how hard tracking and fixing something can be if the regular stuff doesn't do it.

And I have had problems seemingly disappear and then resurface a few weeks later

I do feel that the guys hosting this are working their buts off to fix it...

I do have some concerns about this length of time that are not about them or their competence...but just how servers seem to work

Am I over worrying? I know I am frustrated as I lost several days work on some sites and can't go foward with some other projects and have a few customers (or members of group with sites) who are not happy...Soo I may well be worrying about non existent concerns--thus the post to check it out


Renee

Oh, I thought you were the server owner! :)

They might well be committed to fixing it and that might not be a bad thing, provided they have the proper skills to judge the situation and act accordingly.

Best thing to do is to ask them and talk it out with them. Just from experience, go easy on them when you do. I know how it is to be up for 2 straight days, spending hundreds of dollars to get things running properly and having people yell at you about it. Ain't pretty. So, if you're going to grill them, it's best to do it as nicely as possible.

If you don't like their answers, then it's probably time to start looking elsewhere. If they are attentive and obviously committed to fixing things, then I'd give them the benefit of the doubt. However, that only goes so far.

oh...and if they know what they're doing, they'll get it stable. :)

There is not question they are working on fixing it...

They also say it has now been stable 17 hours...

would you move to a different server or wait it out and assume if the server is fine it is fine?

Before the HD toasted, where there the same performance problems there are now? Presuming it is a new HD -- which should rule out any hardware problems -- the problem is either with the setup of Server or the Server being overloaded.

If the problems happening now were being experienced before, then really, nothing has changed -- as far as the Server performance problem. Shouldn't take that long actually, to figure out what the problem is and correct it.

Being a Reseller you need to determine if it is worth staying where you are or moving on. Standard business decision actually, all types of business have service problems, and you need to be pro-active in these type situations -- or be lucky at making good business decisions. ;)

Originally posted by oracleweb
There is not question they are working on fixing it...

They also say it has now been stable 17 hours...

would you move to a different server or wait it out and assume if the server is fine it is fine?

I'd wait it out and watch closely. Has the box really been stable for 17hrs, or is that just their version of it? Has it been up/down for you or slow?

If they're working on it and you're confident in their abilities, I say stick it out. Though, I'm sure they wouldn't mind moving you to another server to allay your fears. :)