E300

It read like a stressed Indian tech support person.

Issue looks to me to be a failing card on a gateway / router somewhere. Every 2 hours it overloads with errors and fails, secondary card then takes over everything reverts when the original card reboots then repeats x hours later. Very difficult to find the failing card in a large network mush easier in a smaller network BUT fault is never there wen you look!

I haven't opened a ticket yet but I think I will go ahead, just more as a reminder that this is an ongoing issue and not resolved by any means. I'm sure I'll get the usual basic suggestions such as "reboot the router" which won't fix the problem at all.

Following the thread though, it doesn't seem very optimistic given their response to other customers so far. I wouldn't expect this less than stellar quality of support from a business oriented ISP. I'll give them the benefit of the doubt though and see how the next few days progress.

For reference, my BQM used to look something like this for a very long time: https://imgur.com/a/kQJc7i4

I just hope I don't have a repeat scenario like I had with another ISP some time ago, although that issue was related to poor single thread download speed and I was on FTTC at the time. Either way I was in contract and it was a bit of a tussle to get that other ISP to release me from it after countless weeks of no resolution.

EDIT: Used imgur as I was unable to provide a direct link to a historical BQM without it asking the user to login.

Edited by Ixel (Mon 02-Nov-20 14:38:29)

Like other potential Cerberus customers I've been watching this thread with interest to see how Cerberus deal with wide spread network issues like this, Its left me feeling that they still have work to do to meet the expectations of their customers when it comes to issues like this. I hope after they have finally sorted this ongoing issue they will sit down and do a deep dive review and put procedures in place to deal with it better should it ever occur again.


All the best to all those affected.

Seems to have been some change since 7am this morning. Rather than an hour of poor latency followed every couple of hours by things being okay, it's now consistently ropy with no obvious pattern. At least that is what I'm seeing now.

I've also received a response to my ticket saying the network team have said the issue is resolved!
https://www.thinkbroadband.com/broadband/monitoring/...

For context, this is the chart the day before the issues began:
https://www.thinkbroadband.com/broadband/monitoring/...

Looks like in a month I'm an ex customer if this is the new norm.

Edited by E300 (Mon 02-Nov-20 16:13:18)

I've also had this message from their support team at 16:10 today:

"We have double checked with our Network team and as per them the issue is already resolved. So request you to please check and let us know in case of any further issues."

My response was:

"This issue is clearly not resolved. The BQM for my connection and for the gateway @ 46.37.48.1 are attached and I’ve included links to each respective live BQM monitoring session for information.

As you will see there are still regular latency delay spikes occurring. Although the situation has improved from late last week, these delay spike are coinciding, timing wise on the main 46.37.48.1 gateway as well as for the router for my connection. This is clearly not “normal" operation at all - and not good enough.

There is very evidently still a problem somewhere in your network. Therefore please do not close the ticket. I would like it escalated instead and will call tomorrow to get an update."

They are clearing trying to brush this one under the carpet, but I'll not let it die and suggest anyone else affected keeps their tickets open and pushes for a resolution.

In reply to a post by kitcat:

E300

It read like a stressed Indian tech support person.

Issue looks to me to be a failing card on a gateway / router somewhere. Every 2 hours it overloads with errors and fails, secondary card then takes over everything reverts when the original card reboots then repeats x hours later. Very difficult to find the failing card in a large network mush easier in a smaller network BUT fault is never there wen you look!

That seems a very plausible explanation. The periodicity of the fault has changed as they have implemented workarounds or 'fixes' but the regularity is still there, although it appears to be on a longer cycle time right now.

I don't understand how that can claim it's fixed. The evidence is so clear its not.

I'm not sure but it feels like I'm either being strung along or taken on a "fishing expedition" on this...Latest update /request from support:

"Thanks for the reply.

Request you to please monitor Cerberus DNS IPs: 46.37.32.22 or 46.37.33.22 and update the result over the email to check this further.

Also, please note that the latency you are seeing today is much lower than Friday. Also, it’s not following the pattern that we was very clear on Saturday.

Could you please also enable SNMP on your router with the SNMP string and also allow SNMP over the WAN, so that we can also configure our monitoring system for your connection to monitor the latency/bandwidth further for testing purpose."

What do their DNS servers have to do with this? I'm setup to use google DNS in any event.

Edited by Pheasant (Mon 02-Nov-20 18:09:47)

Thanks for persevering.

I have no idea either why they want you to monitor their DNS servers and if they are monitored what does it prove either way? Don't they have their own staff using the service they already monitor and have their own connections in their office they can test?

Yes true, the latency/jitter we are seeing is much lower than Friday, which was extreme latency with a loads of packet loss as well, but it is magnitudes higher than the latency and jitter on Thursday and everyday before then, and certainly worse than you would expect to see on any business oriented ISP connection.

I will keep my ticket open and let them know I'm also still experiencing problems.

Edited by E300 (Mon 02-Nov-20 18:35:39)

The irony is that I realised I'm already monitoring their primary DNS server, as I was asked in my email discussions with them on Saturday to BQM monitor "cache0.cerberus.net.uk" (46.37.32.22) - which as it turns out is their primary DNS...I just didn't put two and two together at the time.

So anyways in the interests of helping them out, I will put a BQM on on their secondary DNS, aka cache1.cerberus.net.uk (46.37.33.22)

Its only fair, that I post an email that I received a short while ago from Bob Hendy, who is the Director at Cerberus:

"The team do not consider this issue resolved and the support case will remain open until you have confirmed that you are happy. I am sorry if we gave an impression to the contrary. There is clearly still an issue albeit impacting some users – although significantly less than when first identified.

The senior engineers are continuing to look at the root cause of the problem. It is not clear to me how many customers are still affected by this but the issue does not appear to be widespread.

Regardless of the number of users affected, the problem is still apparent on your circuit and needs resolving ASAP and we are working to ensure that happens.

Please be assured that we are on the case and will keep you posted."