STR New Tartans - Automatic Updates Down Again

[figheadair]

The National Archives is once again (previous thread - here) having software trouble meaning that Automatic Updates of new designs are not happening.

Details of new designs can be found on the STR's What's New page. This page is now updated weekly and so 'new' tartans are not held in this section for as long as was the case previously.

[Mike_Oettle]

That is a problem I noticed as well, Peter. I dropped Alison Diamond a line last week about it, but have heard nothing back.
Regards,
Mike

[figheadair]

That is a problem I noticed as well, Peter. I dropped Alison Diamond a line last week about it, but have heard nothing back.
Regards,
Mike

Alison has moved on and her role is now carried out by Mark Burgess. I have raised this issue with him.

[figheadair]

A month on and the problem is still not solved! This is an extract of an email recieved by members of the Advisory Commitee this morning.

I apologise for the disruption to this service and I would like to assure you that this is our top priority to resolve. The issues we are addressing have impacted on a number of our web services in addition to the Scottish Register of Tartans and all our efforts have been focused on resolution however this has proved to be challenging to resolve and involves a process of elimination of potential causes. This is taking some time and I am sorry to report that services remain unavailable and will continue to be for the next few days.

Bad news for a Governmental organisation to be offline for such a long period in this day and age.

[Domehead]

I can no longer get the ".gov.uk" home page to open, even.
Ryan

[figheadair]

Update from National Archives today.

Further to my email below I am sorry to report that despite all our efforts during the course of this week that we have been unable to resolve the issues preventing us making the Scottish Register of Tartans web site available. We are making progress in relation to eliminating the causes which stem from a server failure however I can confirm there has been no loss of registration information. I apologise again for the disruption and I can confirm that the site will continue to be unavailable over the next few days.

[ScottishStuart]

The site appears to be back up and running for the moment. The last new tartan was just over a month ago, so I am guessing they are still working on that part.

Good to see it back though!

www.tartanregister.gov.uk

[figheadair]

Yes, up and running again but we, the Advisory Committee are now running a month behind schedule so it will take a while to get through the backlog.

[unixken]

I submit that the Advisory Committee should order the development (and annual testing) of a Disaster Recovery / Business Continuity plan for all their servers. There should be no reason why a working system, assuming backups are done nightly, can not be fully restored/recovered back to it's last working state, following repair/replacement of failed components, or as a "bare metal install" on a complete replacement system.

At the very minimum, the I.T. people should have been able to install Linux and the Apache web engine onto a machine with a public-facing web page saying "The site is down. We are working on it. Please check back later."

If nothing else, use a system with removable hot-swapable mirrored hard drives. Each night, pop one of the hard drives out of the chassis and plug a replacement drive into the open bay. Take the removed drive off site to another building. The next night, you bring the off-site drive back in, remove your spare drive and put the previously off-site drive back into service. If the machine catches fire, you plug the off-site drive into a new server, turn the power on, and WHAM! You are back up and running with last night's running configuration.

Edit: Then do your troubleshooting/diagnostics with the failed configuration offline, with a development server.

Other recommendations:
A.) Ensure proper Change Controls exist to record (and communicate internally) when patches or program changes are made, so that when something breaks, your troubleshooting knows exactly what has changed, and when (and can easily be backed out.)
B.) Enable full logging (even at the O/S level, if your apps themselves don't have verbose logging), to more quickly pinpoint the cause of a failure.

[Pipersson]

Jipeeeh!

And here's the tartan in question...but with a closer look it is ...much different.

http://www.tartanregister.gov.uk/tar....aspx?ref=2710

Tom