[availability] Here we go again. One year later and more capacit
Last year at this time were were seeing a lot of issues with Fairpoint's dsl service. Looks like we're in for another bout of problems this spring.
Things had been working pretty well since last spring but a few weeks ago we started seeing momentary outages of 20 seconds or so. We remained connected to fairpoint but could only get as far as our gateway. Other times we would get as far as their peering point with alter.net. No latency issues or packet loss. Just nothing moving for 20 seconds. This would make a web page slow to load once in a while but was very bad for gaming, streaming and terminal sessions.
On the 16th they experienced a "routing issue" that lasted from 1AM or so until 9 or 10 AM that effected customers with static addresses. Complete outage, nothing moving past the gateway.
Since that time the outages have become more frequent and longer in duration. Now we seen them a half dozen times an hour and they last for 45 to 60 seconds. Same last hops, either the gateway or alter.net.
Called support and was informed that the problem that occurred on the 16th has not been resolved as yet and they have no estimate for resolution. He characterized it as occasional slowness. This might be true from the perspective of web browsing but not from the perspective of streaming, terminal sessions or gaming - for these types protocols they are hard breaks.
Also seeing some last packets now that we weren't before. Its been as high as 10% at times. Latency hasn't been a particular issue.
He took the obligatory notes and opened an up steam ticket and we should get a call after they check things or if they need to come tot he site - hoping they don't put some poor repair guy through that because its pretty obviously not a last mile problem.
Seeing the other threads re fairpoint that are on here right now I would hazard a guess that they are having capacity issues again. You would hope that they would be a bit more proactive and keep capacity ahead of the usage curve by a reasonable margin.
( For last years issues see »Fairpoint and ALTER.NET peering
Re: [availability] Here we go again. One year later and more cap
Yeah, I've been noticing ~20 second drops too,it kills any new sessions from being made and it kills my work VPN.
From what I can tell it looks like a peering between Fairpoint and ALTER.Net again. There is some serious ugly going on which seems to point to a failing/overloaded router. Given that we aren't seeing the huge latency spike I would venture to say it isn't a network congestion issue. This is a hard drop, like an over loaded router is crashing or just seizing up and dropping packets.
Ether way I've been battling Fairpoint for well over a year now trying to fix my own issues. As a network engineer, I don't know what Fairpoint is thinking. Known issues like this should be attacked with a vengeance, especially if it is a problem with one of the core routes. Kill it with fire, nuke it from orbit, whatever it takes; problems like this shouldn't exists for more then a few hours to a few days at most (assuming replacement router needs to be shipped, config-ed, installed).
Progressively worse today.
From 3pm until 7pm it was very broken.
Noticed an odd thing this time, couldn't traceroute to anything even the gateway but I could ping one of their dns servers. I got a response literally every other packet. 50% packet loss.
Now its in and out every few minutes.
We have a static ip. We have family across the street who have a dynamic ip and that seemed to be working while we were out.
Router issues for sure. Though every other drop appears to be at alter.net so peering issue in there as well.
Called again tonight. Squeaky wheel.
Was on hold for a long long while so suspect they were busy.
They haven't found a fix yet but are working on it. When I asked specifically if it was just us or something more wide spread he allowed there were continuing to have issues in Maine state wide with static addresses with no fix in sight.
Probably very clear if you know the details but its hard to imagine a situation where only static addresses are involved in an issue that goes on for this long without them being able to make a fix. The first signs of it started at least a few weeks ago.
I had the same issue in Portsmouth NH yesterday around 5pm, and now this morning again around 11am. Work connection on static IPs has 600+ ms to the gateway and minimal traffic passing through to the greater internet. Some connections stay active but slow, others are dropped and new connections time out.
Home service with a dynamic IP (no PPPOE) had no issues during the same time.
Had to swap over to a backup Comcast connection to let everyone in the building continue working.
Oh and this is with FAST, not DSL.
|reply to dkh |
Fairpoint is terrible and they will probably never fix it unless 100s of people complain. I have their DSL service and it's absolutely horrendous. After years of problems and tech support, they finally admitted that my town is run off 5 T1 lines, and they have no ability to see the current usage on the local equipment, and admitted they have about 35 customers on this, all quotes as having "3mb" speeds.
They just signed up another person in town, so it's obvious they are just going to continue to make it worse by adding more customers without ever upgrading the equipment. They have told me about upgrades for the last 3 years and all they've done so far was add 1 additional t1.
Good luck with your problem, but I doubt they care.