Can someone read this?

**David Copeland** · Tue 13 Mar '01, 11:17pm

48 people have read this and . . . ?

**KeithMcL** · Wed 14 Mar '01, 3:26am

I don't know the answers to q's 1-3 but for #4 I can tell you that I use Netwhistle.com and their service is quite good.

The email you get from them tells you pretty much what was wrong when it last tried accessing your site.

I'm sorry, I don't have a copy of the mails it sends to you right now.

Another service like it is www.serverrat.com

rgds,

**M. Libbert** · Wed 14 Mar '01, 10:15am

Originally posted by David Copeland
48 people have read this and . . . ?

“Silence is golden” my friend

But seriously I can’t answer your question either

**David Copeland** · Wed 14 Mar '01, 10:28am

Here is the latest from my host, which has me worried.
(PLEASE Reply with any of your comments)

When we reported that our forum and web site
went down, we got the following messages:

----------------------------------------------------------

MESSAGE 1 (Tuesday 12:30 PM ET):
"Looks like HTTP was over run. The server is rebooting
right now and should be back online in the next 10-20
minutes."

----------------------------------------------------------

MESSAGE 2: (Tuesday 10:22 PM ET)
"A bad CGI script could cause the server to go if it
doesn't execute or stop executing properly. This is
the most common cause.

"This is not your fault. It's not a single site owner fault,
so to speak. It just happens when a script doesn't work
properly for whatever reason, which happens no
occasion. We are working on improving the server's
kernel and it's ability to make sure these scripts are
dealt with before they take a server offline.

"(The server) stalls and waits for someone to notice it.
It's hard to make a server reboot on it's own (from the
server point of things). You can have scripts and surge
strips that will hard boot (turn power off and than on)
a server if it goes down. Our servers are monitored by
several different monitoring devices internal and
external. If one is not responding, we are paged so we
are notified... but it doesn't automatically reboot.

"Depending on the severity, we can sometimes login
and reboot it nicely, sometimes we have to actually turn
the power off remotely, and sometimes a human
actually has to physically sit there with the server to
clean stuff up (the worst).

"Unfortunatly, this issue is more severe than originally
diagnosed. The server suffered a very bad corrupted
file system on one of the partitions.

"This isn't easy to fix. We currently have a systems
administrator working on the server and they expect
to have it back online within 4 hours.

----------------------------------------------------------

MESSAGE 3: (Wednesday 1:35 AM ET
"Only the /usr partiition was corrupted which holds files
that actually run the Web server. But this directory does
not contain any user data. All of that information will be
in tact 100%. To keep it that way though, does take
longer, hence the delay in not having the server up
quickly."

----------------------------------------------------------

MESSAGE 4: (Wednesday 10:27 AM ET
"At this time, we are working on recovering the data on
the hard drive. This is a tedious process.

"We do not have an ETA yet. Once the hard drive data
has been recovered, it will be about 1 hour to restore
the rest of the server."

----------------------------------------------------------

MESSAGE 4: (Wednesday 1:21 PM ET
"Recovered" was the wrong word. The corrupted
partitiion needs to be repaired and then we will be
pulling all the user data off the drive. The user data
is in a seperate partition on the drive that was
unaffected, but to access the drive, the /usr directory
has to be in tact because it hold the server files.

"The data is being pulled of the drive now. It will take
about an hour,

"Then it will be uploaded to the new server and
booted... should be running within 2-3 hours.

----------------------------------------------------------

Update: It's Wednesday, 4:23 PM ET:
We now have the Apache webserver software up, but
our data has not yet been restored.

Comments? Preventative solutions, if any?

**M. Libbert** · Wed 14 Mar '01, 11:11am

Give them some time to get it back up. If two days pass and no data is restored cut your losses and set up A.S.A.P on another host. I think we both waited too long when this happened to us the last time.

**David Copeland** · Wed 14 Mar '01, 11:20am

Thanks Mark, for the kind reply.

Yes, you and I were on another host last year and
have gone through some horrible times. We've
since gone separate ways on a host these days.

The email that my current host is sending me is
cordial and respectful, but I believe he uses another
provider that he has little control over. His hands
may be tied.

But back to my questions. Would any other host
still have the same challenges? If not, what do
they do different that can prevent this mishap
from happening again?

Anybody?

Thanks,

David

**David Copeland** · Wed 14 Mar '01, 1:17pm

Any help from you guys on advice would be greatly
appreciated.

We're still down.

I'm worried that the repair and the backups are
no problem.

Anyone else have this problem?
How did it go?

David

**chrispadfield** · Wed 14 Mar '01, 1:32pm

i would recommend a new host if it is still down!

**David Copeland** · Wed 14 Mar '01, 1:36pm

That is certainly a consideration. But I'm more concerned
about our data at this point.

Is this something that is a 50/50 chance of recovery?
Or has it been the experience of anyone here that
what our support has described is a cake walk to
full recovery?

**Dark_Wizard** · Thu 15 Mar '01, 2:27pm

Dave's site is up and running on a new host...his data is intact.