OFFICIAL DOWNTIME THREAD

[MrHanky]

Still Down???

[Tigurius]

No, it's been up for a while. It just seems the disconnect issue is back with a vengeance. I've logged on a couple times and only been able to maintain a connection for about 30 seconds at a time - long enough to put a potion or two into my beehives.

Or rather it WAS up when the above posts were made - looks like it's down again for some fixes.

[SnaponTools]

it came back up but not back down lol coli is working on it im sure

[Dramoor]

Man I just finally got power back after a crazy storm last night and then the server is down!

[Albino]

Was that the same storm that roared through Chicago? Cuz that was a wicked one

[Tigurius]

This is kind of behaving like a dying network interface controller. The behavior is very similar to what we used to see when I worked at Cisco whenever we had a defective or failing min-GBIC or interface card. Depending on what kind of connection you have though it could be a component of a Layer 3 switch or your server. Pathping (not the best tool I know) run during connectivity problems shows 0 dropped packets though so I'd guess it's the GBIC or NIC on the server itself rather than your router or layer 3 switch.

Of course I'm assuming you've got the server on a network you own and not in a hosting facility. If that's not the case the problem could be on their end.

[upwind48]

I think there might be a clog in one of the internet tubes.

[Tigurius]

I think there might be a clog in one of the internet tubes.

Thank you internet ghost of Ted Stevens. It's funny how someone who was so clearly unfamiliar with something could leave such a lasting impression on its culture.

[Dramoor]

by Albino on Wed Jun 22, 2011 11:27 am

Was that the same storm that roared through Chicago? Cuz that was a wicked one

Yes Very Same Storm! Villa Park, IL here

[Agony]

This is kind of behaving like a dying network interface controller. The behavior is very similar to what we used to see when I worked at Cisco whenever we had a defective or failing min-GBIC or interface card. Depending on what kind of connection you have though it could be a component of a Layer 3 switch or your server. Pathping (not the best tool I know) run during connectivity problems shows 0 dropped packets though so I'd guess it's the GBIC or NIC on the server itself rather than your router or layer 3 switch.

So basically you pinged the server

[Tigurius]

This is kind of behaving like a dying network interface controller. The behavior is very similar to what we used to see when I worked at Cisco whenever we had a defective or failing min-GBIC or interface card. Depending on what kind of connection you have though it could be a component of a Layer 3 switch or your server. Pathping (not the best tool I know) run during connectivity problems shows 0 dropped packets though so I'd guess it's the GBIC or NIC on the server itself rather than your router or layer 3 switch.

So basically you pinged the server

Nope.

Ran path-ping for 10 minutes when instability started and into a reboot of the server. Since all ping attempts during that period succeeded (even while the server was rebooting) I am fairly confident I was not pinging the server at all but rather a router, layer 3 switch, or firewall which was the last point in the path which was configured to respond to pings. So yeah that really only suggests the problem exists after that last point.

Also the connectivity issue had been getting progressively worse since people first started complaining about it a day or two ago - at which point many people had not even experienced disconnects yet. That and the fact that it appears to work fine for a few minutes after reboot and then becomes almost impossible to stay connected for more than a few seconds is what reminded me of dying NICs and GBICs. When we had those their failure appeared to be time and/or load based and grew progressively worse until the connector or card or other component entirely failed. We also saw somewhat similar behavior in dying and overheating layer 3 switches.

[Agony]

Actually yeah, the dying NIC issue is what I noticed also. We get the same issue here sometimes

[Golden Temptress]

Funny my internet crashed just before the server went down, alot of internet systems are going wrong at the moment

[+Colibri]

I'm in the process of some rewiring. I don't want to give any ETAs but i'd estimate it at about an hour till the shard is back under the rewired configuration.

Well, could be 15 minutes but i want to make sure it's all working ok... rather a bit more downtime than for there to be disconnects again.

[Riath]

rather a bit more downtime than for there to be disconnects again.

Anything worth doing, is worth doing right. Cheers for doing the right thing, rather than band-aiding it and hoping for the best.