Forum Down

[thescruff]

I would have thought Root or BillyGoat could throw some light on this as it clearly an inhouse problem with one or all of the 3 racks.

[gas4you]

This is the current results from here with everything flying


1,192.168.2.1,2ms,None,----
2,212.74.111.217,29ms,lo99.lon-1-sip.as91,----
3,80.40.119.145,30ms,ge1-0-0-23.lon10.as,----
4,10.72.11.73,45ms,None,----
5,80.40.155.37,30ms,xe-9-3-0.bragg001.l,----
6,80.40.155.62,30ms,xe-7-0-0.scr001.loh,----
7,78.144.0.190,29ms,host-78-144-0-190.a,----
8,195.66.225.138,31ms,linx.redstation.net,----
9,109.200.16.217,36ms,217-16-200-109.rack,----
10,109.200.16.186,36ms,186-16-200-109.rack,----
11,109.200.27.10,35ms,10-27-200-109.rackc,----

[thescruff]

Why has Dave got different IP addresses for the Racks, to Wine~o and me.

Could that be why some have the problem when others don't

[wine~o]

Why has Dave got different IP addresses for the Racks, to Wine~o and me.

Could that be why some have the problem when others don't

Me neither ...Don't even know what a rack is....

[thescruff]

Rackcentre what UHM connects too.

One of the three was down for nearly an hour, I wondered if one or two went down whether it put load on the remaining servers.

[ultimatehandyman]

Support are still looking into the issue

[thescruff]

Support have answered and have done the following-

I have reconfigured the connection limit and settings of apache now to allow for more traffic and better handling and flushed the firewall. I also set a cron to restart apache every 6 hours to keep it refreshed and avoid any build up to cause it to lag.

Hopefully this will eradicate the slow loading page problem

Whatever they did here emulated the fault perfectly.

[thescruff]

13 109.200.16.206, 206-16-200-109.rackcentre.redstation.net.uk timeout

Timed out for the last 40mins just back.

[thescruff]

Slowing up a tad.

[wine~o]

Slowed right down again for a few minutes...


1,192.168.1.1,1ms,None,----
2,----,Timeout,n/a,----
3,78.151.225.209,18ms,host-78-151-225-209,----
4,78.151.225.236,19ms,host-78-151-225-236,----
5,62.24.240.13,20ms,xe-11-2-0-rt001.the,----
6,78.144.1.4,19ms,xe-11-2-0-scr001.th,----
7,78.144.0.160,21ms,host-78-144-0-160.a,----
8,195.66.225.138,20ms,linx.redstation.net,----
9,109.200.16.217,24ms,217-16-200-109.rack,----
10,----,Timeout,n/a,----
11,109.200.27.10,25ms,10-27-200-109.rackc,----

[root]

a traceroute will map your particular isp's route to an end point so for different isp's/people the hops will be different depending on how complicated the routing is to the destination. to be honest without gathering data from one source to the destination and then being in the datacentre where i can pinpoint the systems that are routing the traffic into the racks I wouldn't be able to tell you much. Although if you restart apache using a cron job like apachectl | graceful, this will more than likely destroy the session variables on the server causing people to have to re-log in sometimes.

[root]

which ISP's are you guys on?

[wine~o]

Talk talk, though don't think that's the problem as different users using different Isps are having the same problems, though sometimes some are whilst some are not...if that makes sense..

[root]

yup makes sense, could be loads of things causing it, a load balancer in the datacentre could be out meaning when people are routed through that machine they dont get anywhere.... a dns server somewhere en-route from isp to datacentre could be malfunctioning and taking a long time to resolve the address.

[thescruff]

With Virgin.

If it was a datacentre en-route from isp, why would a range of users have the same problem.

Would one rack going down put extra traffic on the other two.