kelvinq

Pass78 goes down again. 5 (was 4) out of my 7 websites are inaccessible.

Websites down include - kquee.com, quee.org, interesthink.org, sentigage.com and INTERESTHINK.com

Other threads reporting the same problem are -

Seriously, pass78 down again!
Pass78 Down?
[Answered] pass78 problem again

If this problem keep recurring, should Surpass be looking at either moving some resellers to another server OR capping the number of sites hosted on each reseller account?

Surpass - I think explanations are no longer sufficient. We want solutions from you. How are you going to prevent this from happening again?

(That's assuming that we are somehow going to tide over this episode.)

Kayla

For the downtime in this moment, moving resellers or capping is not the issue.

The issue is one wildly abusive script uploaded by a reseller's customer. Seems it went haywire on March 15th as well. I've disabled it.

__________________

kelvinq

Is it an issue with a wild script or is it an issue with Bind?

Because Mike says that it's a DNS issue -

Pass78 (66.7.212.189) : Resellers

kelvinq

UPDATE!

All my websites with Pass78 are down. Whee!

Mike

It is an issue with both. I will update you soon

__________________
Mike
Surpass Special Operations

Mike

Kelvin and pass78 clients-
Yesterday upon first glance and notification when checking pass78 I immediately noticed a problem with dns service. My main concern was restoring service and I neglected to look through the servers cpu/memory history. As I explained in a previous post related to pass78 inaccessibility, cpu/memory resource abuse violations are possible and frequent in the reseller shared environment, this should have been checked. Kayla picked up and noticed this thread and found the culprit that she provided a SS of the high CPU usage. This clients usage also appeared high on 3/15/08, they are now suspended. Moving forward from the issues on this server in the past 7 days I will add an high load reporter as a tertiary notification and monitoring method to further preserve the uptime on this server.


Server Status for HDU434 pass78.dizinc.com
Service Alert Age Status Check Result
http 16h 55m Silence Up Connection established (0.0177 secs) Received: HTTP/1.1 200 OK
pop3 6h 60m Silence Up Connection established (0.002562 secs) Received: +OK Hello there.
ssh 16h 56m Silence Up Connection established (0.015963 secs) Received: SSH-2.0-OpenSSH_3.9p
smtp 3h 46m Silence Up Connection established (0.103352 secs) Received: 220-pass78.dizinc.com ESMTP Exim 4.68 #1 Wed, 19 Ma
ftp 6h 60m Silence Up Connection established (0.007672 secs) Received: 220---------- Welcome to Pure-FTPd [TLS] ----------
imap 7h 18m Silence Up Connection established (0.018489 secs) Received: * OK [CAPABILITY IMAP4rev1 UIDPLUS CHILDREN NAMESPA

Server Reboots
Time Submitted Time Closed Technician Cause
Details 03/14/08 - 12:57 PM 03/14/08 - 01:03 PM Rafael Ramirez Does not require reboot
Details 02/13/08 - 05:35 PM 02/13/08 - 05:58 PM Rafael Ramirez Server Overload

07:26:29 up 16:54, 4 users, load average: 0.41, 0.64, 0.66

__________________
Mike
Surpass Special Operations

kelvinq

Hi Mike,

Thank you for the detailed explanation.

> she provided a SS of the high CPU usage.

What's "SS"?

I think the load monitor is a great idea. Prevention is better than cure!

Thanks again Mike for coming back to us.

Kelvin

Mike

SS= screenshot
The load monitor I installed is something I will personally use with our support teams to keep an extra eye on this server. This is something small in addition to our already proactive monitoring.

__________________
Mike
Surpass Special Operations

Dannycg

Thanks Mike, for all the work on Pass78 it's greatly appreciated

__________________
Pass78