[resolved] OUTAGE: broadband lines are down for some customers

This was resolved approx 4pm after the faulty switch was swapped on the supplier network

[Update at 14:46]

The supplier is advising us that most lines are now returning on-stream; this may take a few more minutes as the Radius catches up. Anyone still affected after this time should power-off their router for at least 20 minutes to clear any stale session. Please email into support@merula.net if this fails to bring you back live.

We apologise for the lengthy downtime and are looking at further remedial work with the supplier to ensure that such a failure doesn’t affect us in future.

[Update at 14:33]

Apologies for the lack so far of anything concrete in time-to-fix terms; we are escalating this to senior managers inside the supplier to get this fixed.

[Update at 13:33]

Senior engineers are currently on site working on the faulty hardware.

Further updates will be posted once the work has been completed.

[Update at 12:35]

Supplier update: We’re seeing a partial recovery on the equipment.

We’re aware some circuits are still down, our engineers are looking to replace some of the
hardware in the switch stack.

Further updates will be posted when available.

[Update at 12:10]

The supplier has a new switch on route to the site to be swapped out — they’re expecting this to complete by 1pm. We’ll update as this progresses.

We are aware of a problem affecting one of the interconnect switches on a transit supplier network which means that a number of lines dropped earlier this AM and are still down; they and we are working on getting this switch bypassed and replaced. Currently we have no time-frame for a fix but believe this will not be service affecting for too long.

Name Server Upgrade [01/05/2017]

We are upgrading the hardware and software for one of our name servers in Harbour Exchange Square London during the afternoon / evening of May 1st. While this will not directly affect customers – they may find some name resolution is show while the old server is removed and the new hardware installed. The new server pre-configured and should work identically once installed. However as with all upgrades there is a risk

We will upgrade this post once the upgrade is complete or if any issues occur

This work was completed. Further work will be completed on our second name server in Telehouse North in the coming days

UPDATE: Merula support contact details

We have slightly changed some of the detail on our support contact page and for ease of reference append this below:

This site is designed to keep you updated on both planned work and any network outages or service(s) issues at Merula or with our suppliers.

The main support number is 0845 330 0666 (geographical 01480 355566)

There’s a second, fallback (geographical) number: 01480 411616, which rings directly at our support centre. In the event that the phone system here (and the primary support number which routes via here) is unavailable for whatever reason this should be used.

We also suggest that all customers subscribe to our mailing list ensuring that copies of these status messages and updates are delivered to you by email or alternatively if you prefer, to your favourite RSS reader.

Support questions can also be emailed to support@merula.net to raise a ticket.

INFO: logging & reporting issues 4/4/17

We are seeing a few ADSL/FTTC lines that we are unable to monitor internally to track performance and transfer speeds. To overcome this we may need to reboot an internal switch that will mean that the PPP sessions on these few lines will momentarily drop. The circuits should all come back on-line automatically, but if you find an issue please re-boot your router to establish the session.

INFO: mail-server upgrade 4th April 2017

One of our clustered mail-server machines will be updated and moved to a new location at some point tonight. This will not involve any down-time but mail delivery may be slightly slower than usual for a period of up to 2 hours as the databases sync back.

INFO: fibre ring maintenance 7th April

 

This work will upgrade the capacity on one leg of our dark-fibre ring. Whilst it won’t cause any outages, we are at elevated risk during the maintenance window and there may be brief transient routing issues when the fibre is brought back on-stream.

Maintenance Window: 21:00 – 07:00 GMT

Primary Date: 07-Apr 2017

2nd Primary Date: 08-Apr 2017

3rd Primary Date: 09-Apr 2017

Location of Maintenance: London

Reason for Maintenance: supplier will perform fiber maintenance to add capacity

INFO: supplier equipment upgrades 23/24 Feb

We don’t expect this to materially affect any of our services but there may be a few instances of slight packet loss or routing transit issues during this window.

Start: 2017-02-23 00:00:00 GMT (UTC + 0)
End: 2017-02-23 04:00:00 GMT (UTC + 0)

LAN(s) AFFECTED
London Juniper

INTRODUCTION
This maintenance is scheduled for the 2nd stage of Juniper software and Routing Engine upgrades.
During this maintenance we will replace both 32 bit RE-S-2000 Routing Engines in Edge1,
Edge2 and Edge3 THN (Telehouse North) with 64 bit RE-S-1800×4 Routing Engines. We will also upgrade the Junos version to 15.1F6-S3 from 13.3R5-S4.1.
All members connected to edge1-thn, edge2-thn and edge3-thn shall experience downtime of up to 20 minutes whilst the RE’s are swapped and rebooted.

INFO: Netgear router vulnerabilities

The following routers are vulnerable as below. New firmware has been released for some of the models and is available from here: http://kb.netgear.com/000036386/CVE-2016-582384

Vulnerability Note VU#582384

Netgear R7000, firmware version 1.0.7.2_1.1.93 and possibly earlier, R6400, firmware version 1.0.1.12_1.0.11 and possibly earlier, and R8000, firmware version 1.0.3.4_1.1.2 and possibly earlier, contain an arbitrary command injection vulnerability.

By convincing a user to visit a specially crafted web site, a remote, unauthenticated attacker may execute arbitrary commands with root privileges on affected routers. An unauthenticated, LAN-based attacker may do the same by issuing a direct request, e.g. by visiting:

http:///cgi-bin/;COMMAND

An exploit demonstrating these vulnerabilities has been publicly disclosed.

Netgear’s advisory confirms that the R6200, R6400, R6700, R7000, R7100LG, R7300, R7900, and R8000 are vulnerable, though affected firmware versions are not enumerated. The vendor has indicated that their advisory will be updated as firmware updates are released.

Possible LNS Issue 01/12/2016

We are seeing some oddness on one of our LNS routers which terminates some of our broadband circuits. We believe this is only affecting our management access and not the service provided over this.

We are working with the vendor to understand this and hopefully resolve this. Assuming this is not customer affecting we may need to reboot this LNS later this evening causing a brief network blip as the sessions re-connect.

IF we find this is affecting service earlier we may have to perform the reboot earlier. We will update this as we know more