Short network disruptions tonight (specific services)

Expected resolution: 29 Mar 2016, 22:00 UTC
Return to issues
Issue status: Resolved Date:

29 Mar 2016
22:28 UTC

Posted by:

Tom Hill

With no alerts raised, and no tickets from any of the affected customers, I'm going to mark this as resolved. We'll still be around for another 30 minutes or so, but please do use the urgent-support address for anything critical after that time.

Tom

Issue status: Investigating Date:

29 Mar 2016
21:54 UTC

Posted by:

Tom Hill

Phase #2 is now complete, and was completed with minimal downtime - MACs were being re-learnt in under 5 seconds.

Contrary to Phase #1, this part appears to have gone quite well. I'll be sticking around to keep an eye on this for some time, however, so if you do have any issues please do not hesitate to get in touch via the support email addres(ses).

Issue status: Investigating Date:

29 Mar 2016
21:37 UTC

Posted by:

Tom Hill

Phase #1 finally complete; any and all public-facing Bytemark services that were affected, will now be universally available from all corners of our network.

Now on to Phase #2.

Issue status: Investigating Date:

29 Mar 2016
21:00 UTC

Posted by:

Tom Hill

Phase #1 is going to overrun slightly, but we should still have plenty of time to complete Phase #2 from 22:30 to 23:00.

Our apologies for the delay; this is simply due to some unsightly spanners appearing during the switchover of Phase #1.

Issue status: Investigating Date:

29 Mar 2016
19:48 UTC

Posted by:

Tom Hill

Quick reminder that we'll be starting this work shortly.

Issue status: Investigating Date:

29 Mar 2016
16:30 UTC

Posted by:

Tom Hill

Hello,

We'll be disrupting the connectivity for some VLANs this evening. This will be done in two stages.

Stage #1, 21:00: We'll be moving some of our own VLAN's gateways to new firewalls. This might impact a few lesser-used customer facing services, such as upload.ns and WSUS services. The downtime should be no more than 15 minutes, and likely much less.

Stage #2, 22:00: Customers with very particular VLAN services will have been emailed about a short (up to 60s) outage for some of their hosts. Anyone to be affected by this work will have been emailed about this - from our ticketing system - a week in advance. The hope is that we can get this outage down to well under 10 seconds, meaning it should only pose a problem for UDP/ICMP traffic.

To summarise: if you've not been emailed don't worry about Stage 2, and Stage 1 should go mostly unnoticed from the outside world.

Many thanks.

Return to issues

Issue still not addressed? Please contact support.