I had my connection drop due to a local power problem, up to then had spent tens of days on the more reliable x.witless, but connected back up to one of the new gormless LNS's which then crashed a few days later. I was bit annoyed not being on x.witless anymore as that seemed pretty stable, alas it seems it just had a good run.

There doesn't seem to have been any updates for a while about the debug logs they've captured and if they are nearer to a fix, perhaps no news means they are no nearer.

I wonder if they have a plan B, for example buying in some alternative hardware?

They say on the service status page that x.witless likely crashed due to not having the newer software and a NVMe drive fitted, yet ironically its stayed up for a very long time, and LNS's having been already upgraded with an NVMe drive fitted have reached nowhere near that length of uptime before crashing. So fitting an NVMe drive doesn't appear to have improved stability that I can see.

Edited by E300 (Tue 27-Feb-24 13:27:15)

In reply to a post by E300:

I wonder if they have a plan B, for example buying in some alternative hardware?

There's plenty of COTS hardware capable of doing this job, but switching to some other make of core router would be the worst possible advertising for the Firebricks. That may be why they're still persisting with them.

A dual-vendor strategy could be wise. Along with an opt-in beta testers group.

Some overnight work tonight (early hours of 1st March) so some drops and shuffling about again, they are also separating out CityFibre and BT/TalkTalk customers so we are on separate LNS's and don't mix

They've not said if they've found a problem and the software update is a fix for the drops, still it's not easy trying to fix something you can't replicate at will.

https://aastatus.net/42630

Was with AAISP for 3 months via City fibre and I can honestly say it was the most unreliable expensive over rated ISP I have ever been with. I changed over to IDNet and I have had absolutely no problems at all plus I'm saving money happy days.

In reply to a post by jalzoo:

Was with AAISP for 3 months via City fibre and I can honestly say it was the most unreliable expensive over rated ISP I have ever been with. I changed over to IDNet and I have had absolutely no problems at all plus I'm saving money happy days.

It might be cheaper and more stable but you don't have Continuous Quality Monitoring anymore.

In reply to a post by jalzoo:

It might be cheaper and more stable but you don't have Continuous Quality Monitoring anymore.

I'd take a £13 a month reduction & a stable connection over quality monitoring any day of the week. If I'm really that interested in monitoring the quality of my connection I would just use the free ones out there..

Edited by jalzoo (Wed 06-Mar-24 15:11:13)

Still not fixed...

https://aastatus.net/42636
Customers on the X.Witless LNS dropped and reconnected at 11:35 today.

Indeed not. However they have split off City Fibre connections from the rest. I'm on BTW and currently connected to g.gormless. So possibly BTW/TTB connections are on the 4 gormless LNS with City Fibre on the 3 witless lns. Further conjecture is the thought that maybe the issue is triggered by CF. Looking at the history, the issue does appear to have started afer taking on CF. Just a thought.

In reply to a post by bellerby:

Indeed not. However they have split off City Fibre connections from the rest. I'm on BTW and currently connected to g.gormless. So possibly BTW/TTB connections are on the 4 gormless LNS with City Fibre on the 3 witless lns. Further conjecture is the thought that maybe the issue is triggered by CF. Looking at the history, the issue does appear to have started afer taking on CF. Just a thought.

I did wonder as well if they thought the City Fibre traffic was somehow causing a bug, hence the separation of traffic. One thing seems clear, given X.Witless had the longest uptime out of all of them without an NVMe drive, and with an NVMe drive it has crashed with a very short uptime and all the others appear no more stable with an NVMe drive either, suggests that isn't playing a part in the stability, or, it fixes a bug they've seen with artificial load testing, but that isn't the same bug causing the issues on live.