Portal Home > Knowledgebase > Industry Announcements > Web Hosting Main Forums > Providers and Network Outages and Updates > TekTonic Atlanta Issues
TekTonic Atlanta Issues
| Posted by TheWiseOne, 02-05-2008, 07:19 PM |
|
It looks like GNAX tripped over our main network uplink or something so things are down there until they can figure out what they did. I will update when it's resolved.
|
| Posted by TheWiseOne, 02-05-2008, 07:37 PM |
|
Ok, now it looks like they have are having some sort of network issue. Our gateway IP 1 hop from them is experiencing 90%+ packet loss.
|
| Posted by webcs, 02-05-2008, 07:48 PM |
|
Oh here we go again, a major outage and no way to use their suppor ticket system. Has anyone been able to reach them.
|
| Posted by TheWiseOne, 02-05-2008, 07:56 PM |
|
Call 484-948-1124 and press 2 to talk to a support rep.
|
| Posted by TheWiseOne, 02-05-2008, 07:59 PM |
|
TekTonic Live Chat: https://server.iad.liveperson.net/hc...ges/liveperson
Last edited by TheWiseOne; 02-05-2008 at 08:05 PM.
|
| Posted by octeto, 02-05-2008, 08:17 PM |
|
The chat works fine. At least we have that.
sadly this happens for the second time in a year, lets hope this one takes minutes and not a day...
regards
|
| Posted by sailor, 02-05-2008, 08:39 PM |
|
matt come on man - that is not true. your port is flapping for some reason and we are troubleshooting it.
|
| Posted by TheWiseOne, 02-05-2008, 08:50 PM |
|
It was coincidental that this issue started happening moments after a tech was working on my racks.
|
| Posted by octeto, 02-05-2008, 08:57 PM |
|
Are we talking of how many hours this time? we are reaching 2 hours downtime already. Happilly enough its not a high demand hour, but we sure hope this doesnt last til tomorrow.
|
| Posted by TheWiseOne, 02-05-2008, 09:16 PM |
|
At this point I think I have a resolution, but I am waiting on GNAX remote hands to do the actual work.
What happened is port flapping between GNAX and our switch started happening randomly -- the reason I stated someone must have tripped over a cable is because this started moments after one of their techs was doing work on a server on one of our racks. Coincidence? Maybe. All I know is everything was working flawlessly and our switch still technically works (it didn't die).
As I said, I do have a resolution in sight, it depends on how fast they move on it.
UPDATE: I do think they passed the blame off on to my switch a little early on this. Something has to be going on with the router I connect to. Here is a ping to our upstream router from another location:
UPDATE #2: For those wondering... I put in the hands on request over 30 minutes ago just to swap where our uplink connects into. Amazing response time.
Last edited by TheWiseOne; 02-05-2008 at 09:23 PM.
|
| Posted by TheWiseOne, 02-05-2008, 09:58 PM |
|
Still waiting...
|
| Posted by octeto, 02-05-2008, 09:59 PM |
|
Seriously....
I dont care whos to blame, lets hope you make ammends and fix whatever is going on, lets be proactive people... i am at least, i still have faith in tektonics good will on providing me service, and therefore i trust gnax on keeping the network going.
So please work together and bring up our servers.....
good night
|
| Posted by TheWiseOne, 02-05-2008, 10:07 PM |
|
I called up Jeff and he says they are working on it. I am going to trust him on his word.
|
| Posted by TheWiseOne, 02-05-2008, 10:38 PM |
|
I was able to speak to a network engineer over there and we appear to be making progress. Assuming no major issues my ETA is going to be 30-60 minutes.
|
| Posted by webcs, 02-05-2008, 10:46 PM |
|
Guys why not move the ticket system to billing.tektonic? That one is ALWAYS up during these problems. Or have an emergency ticket system there. thanks
|
| Posted by TheWiseOne, 02-05-2008, 10:50 PM |
|
The ticket system will be moved to support.tektonic.net and hosted in our Dallas location once HSPComplete supports Cerberus 4. The infrastructure for this already in place, but HSPComplete is not expected to support Cerberus 4 for a few more months.
|
| Posted by webcs, 02-05-2008, 11:01 PM |
|
Whatever it takes to make sure the ticket system is seperate from the majority of the servers if good, thanks...
|
| Posted by Shaw Networks, 02-05-2008, 11:31 PM |
|
It's not GNAX's fault here?
|
| Posted by wbhtlogue, 02-05-2008, 11:47 PM |
|
My VPS is now reachable.
|
| Posted by TheWiseOne, 02-05-2008, 11:48 PM |
|
Everything is working again. What needed to be changed was my port was set to 100Mbit / Full-Duplex on GNAX's router and when their network engineer changed the setting to auto-detection it started working. Why it has worked like this for years and randomly decided not to I cannot say. Both their network engineer and I were completely baffled that this is what ended up fixing the issue.
I'm unsure how to prevent something such as this in the future, except with this experience it's now something that will be checked/changed very early on.
|
| Posted by indigoblu, 02-06-2008, 12:01 AM |
|
ask gnax for another drop for failover? Anything less is another ticking timebomb as demonstrated.
|
| Posted by TheWiseOne, 02-06-2008, 12:12 AM |
|
GNAX won't provide redundant drops. I've fought with Jeff over this a few months ago. I spoke with their network engineer about getting 2 drops from the same GNAX router and he said it should be possible so I'll bring that up with Jeff tomorrow. Although, if something like this happens randomly it could easily happen to both ports too, but yes it would decrease the chance by a little bit. For those that are curious (and because I know it will be brought up), at our Dallas(Colo4Dallas) and Philly(InterNAP) locations we have full redundant network drops into a set of routers running VRRP. At GNAX since we couldn't get redundant network drops we decided to just remove our router and use their router as our gateway, thereby eliminating the single point of failure on our side of things. I'm not really blaming them for this as it's a completely freak event as everything was working just fine with it's current settings and then went kerplunk -- there was no hardware failure on either side.
|
| Posted by octeto, 02-06-2008, 06:54 AM |
|
good idea webcs
And while you do it include a couple of free months service for all the problems with GNAX...
thank god its up now...
|
| Posted by sailor, 02-06-2008, 09:59 AM |
|
again this is simply not the whole truth. we provide redundant drops now for many customers. your equipment or your decision not to run bgp has been the issue to this point. we are in the process of rolling out hsrp and you are on the ticket as one of the first ones to test it.
you have been told this as well. not sure why you are not telling the whole story. but then when we give you hsrp and your single point of failure is your switch - and it dies or the port dies - then what? are you going to go to redundant heartbeat nics in your servers and dual routers?
that is really what you need to do if you are serious about this. hsrp will only help you with 33% of the issue.
|
| Posted by sailor, 02-06-2008, 10:02 AM |
|
we had not changed anything - our standard in the colo is to set the port to full duplex force up.
this is on all ports. however, we did chang our side of the port to accomodate whatever was going on with your switch but we had not made any changes on this switch or port before that. we have hundreds of other connections working in this manner just fine and as you say - its quite baffling - what was the setting on your side in the switch?
|
| Posted by bear, 02-08-2008, 11:03 AM |
|
Closed by OP's request.
|
Add to Favourites 
Print this Article
Also Read
Shared Hosting
Shared Hosting
