Thursday, December 31, 2009

Millennium Bug memories

It's quite amazing to think that it was ten years ago today that we were all getting ready for the Millennium Bug to hit, watching news reports coming in from Tonga, my native New Zealand and other places ahead of the UK in terms of time to see what had (or, as it turned out to be the case, not) broken.

At the time I was working as the webmaster for Telewest, responsible for building and deploying our web server estate along with the sites that ran on them (everything from web design to web server configuration), and it's fair to say that the Bug came as a distraction.

In December 1999 we were gearing up for the launch of the UK's first commercial broadband service - blueyonder - which we went to market with on March 31, 2000 (we brought it forward a day just in case!) and were doing the final testing of all the systems that would support it, as well as building servers and websites. The first version of the blueyonder website can be found here thanks to the Wayback Machine:

I'd also just moved house and got my superfast (not) ISDN line installed so I could work from home, so it really was a completely manic time.

Come the night in question we had a whole team of engineers and other tecchies on call in case of problems, and I remember spending the millennium moment itself logged into our Apache servers checking configuration and refreshing web pages to check they were still working.

On reflection, we spent a lot of time and money upgrading systems in the latter half of 1999 and that probably contributed to the event itself being smooth along with all the preparation for the day - when we also had plenty of additional helpdesk staff in should customers have been experiencing problems.

One man though went well beyond all others in his confidence that there would be no impact - my colleague (and, at the time, boss) Fergal Butler, who decided to spend the millennium holidaying in Cuba - one place where missiles were bound to launch against if the US defence systems did have a problem!

Thursday, December 24, 2009

Google Phase 3

Having now launched Google e-mail for new Virgin Media customers and successfully migrated the whole National base with e-mail addresses over to the Google platform (which has generally gone very well), the e-mail unification team are now onto the next phase of the project – moving the ex-Telewest cable customers with e-mail addresses over to the Google platform.

As we're now in the Christmas/New Year change freeze, the migrations are set to start as soon as we’re back in the New Year - and in the meantime we've completed the staff trial ahead of the start of the migrations.

We found a couple of minor issues that the team are fixing but as with the National staff trial things went very well and the overall feedback summary is probably the most interesting of the results:

As you’d expect not everybody loves change, but generally it's all gone very well and we've closed the trial as a success.

Once the ex-Telewest migrations are completed the team are then full speed ahead to complete the ex-NTL ones (customers with e-mail addresses) and to launch the ‘domain change’, which lets customers adopt a address from one of the old style ones if they wish to.

So it's going to be a busy start to 2010! I hope everyone has a wonderful Christmas and a prosperous New Year.

Sunday, December 20, 2009

Spam survey

Earlier in the year we were invited by ENISA (the European Network and Information Security Agency) - an agency of the European Union - to take part in their 2009 spam survey, where they were surveying ISPs and other companies across the continent about spam, what is done to stop it and the effect it has on customers and providers alike.

We were happy to do so, given the measures we have already deployed across our various platforms to counter spam.

The measures vary at this point in time as we're in a transitional period, with some customers on the Google platform (those with and e-mail addresses), some on the Openwave platform ( and some on the Exim/Microsoft platform ( but all are unifying onto the same technology as we move our e-mail provision to Google, which completes next year.

We've spent quite a lot on the measures in place on our own network, and do drop a heck of a lot of e-mail as spam and viruses - well over 9 out of 10 e-mails usually don't get into customer inboxes as they are not genuine mails.

This is very much a pattern common across the continent going by the survey results:The survey reports only 5% of e-mail gets as far as user inboxes, showing both how effective anti spam measures there are and just how much spam there is. Imagine if your mailbox has 20 times the mails already in it if anti spam measures were turned off!

The results of the study can be found here if anyone fancies a read.

Thursday, December 10, 2009

Hectic Times

It's been a really full on couple of weeks as everyone tries to close down the things they need to get done before they go on their Christmas breaks.

Personally I've got this week off instead, which is giving me a good chance to recharge my batteries.

Our staff trial is underway for the next phase of the Google migrations, with customers having e-mail addresses being the next to move over to the Google Apps platform.

As we're about to hit the annual change freeze, we'll start moving customers over from early January and all going well we'll be onto moving customers with addresses onto Google in no time - but we have a few data cleanse tasks to do first, which are happening in parallel with the blueyonder trial and migrations.

The good news is that customers we have already moved over to Google (with e-mail addresses) continue to track well with regards to key measures like faults and customer satisfaction, and we'll be doing some further research to talk to some of these customers (likely early in the new year now) to see what we can learn to make the migration process even smoother for the next customers to be moved to Google.

Over the last few weeks I've spent a day with our customer service guys in Manchester, including meeting our wonderful Twitter team for the first time:Pete, Billy, Sam and Gareth are doing great work helping our customers out over at @virginmedia.

While in Manchester the broadband product reliability team (i.e. the wider cross-functional group working on service improvement initiatives) got a lot of call monitoring done to pick up some nuggets of things to look into and we also held an intensive 2 day workshop with our technical support guys in Liverpool to look at connectivity faults and how we can improve things like toolsets - which is set to keep me out of trouble for the rest of the year and much of the early part of 2010 I expect!

There's some detailed analysis going on to look into sample accounts reporting connectivity problems and see what we can do to prevent them happening - although when you consider that something like half of all connectivity problems are down to PCs (software & hardware) it's a bit more of a challenge, but one that we're happy to take on.

If I don't get a chance to blog earlier, I'd like to wish compliments of the season to all.