RSS as an administration tool

RSS is obviously a hot topic at the moment and a large part of this comes from the use of blogs by individuals and companies wanting to promote themselves, their products and their ideas.

At the end of last week there have been a rash of announcements and news stories about using RSS in the enterprise. RSS is seen as an ideal way of communicating internally. Thinking about it this makes a lot of sense.

Look at any typical organization and a large amount of the internal email traffic is going to be press release and ‘running commentary’ style information between departments. Numerous people get copied on the latest piece of progress from the sales department; the engineering department’s new lathe installation, or the IT departments news and announcements of services, upcoming maintenance schedules and other information. Time management experts like David Allen will tell you this a contributor to information overload.

For example, NewsGator is creating an enterprise service (Dino) that can be installed behind the corporate firewall to provide an interface to internal RSS feeds through their web-based interface, or through the Outlook connector that enables RSS reading while accessing your email.

For the generation of the RSS feeds Nooked provides a service specially aimed at creating enterprise level feeds to the public for corporate communications, marketing and public relations. Alternatively, you may want to use something like KnowNow 3 Enterprise Syndication Server (see also KnowNow offers Enterprise RSS. This is a middleware product that pulls information out of your databases and in-house systems to generate RSS feeds for both internal and external consumption.

What I haven’t seen – and I’m most surprised at - is a lot of talk about the use of blogs and RSS as a technology to aid IT administrators. When introduced as a part of a wider set of tools and systems RSS (and blogs) can be a very effective way to help run and track information in your IT department.

I use RSS and blogs in combination with a suite of custom administration and monitoring scripts to manage all of my systems, which stretch from a heterogeneous network right through to co-located and dedicated servers at the ISP. They all communicate through a set of servers that are designed to monitor everything. There are three servers, but only one is need, the others act as a hot backups in the event of failure of the main monitoring machine which would obviously defeat the entire process.

At the simple end, logs generated by various components are recorded permanently, with a summary entry generated that is then provided as an RSS feed. Another component monitors the logs, systems and services and picks out anything that it thinks is important. For example, a service going down, lack of disk space or seriously overloaded system, and publishes that to another, separate, RSS feed. Now everybody in the IT department – and any other interested parties – can subscribe to the feeds and keep an eye on the systems as part of their blog reading activities.

Finally, I have an internal blog set up whose sole purpose is to record activity on the machines and configuration in the network. It’s basically one big changelog; if I update a system, reconfigure a service or fit new hardware, I make an entry in the blog. This has the advantage that I log not only what I did, but also when I did it. Then when something goes wrong or no longer works the way it should, I can feed back through the blog to see if one of the changes is the root cause of the problem. The blog generates it’s own RSS feed (it’s based on WordPress) and so other members of the IT department can see what has changed too either by reading the blog, or by subscribing to it.

I have some mixed feelings about where the blog is hosted. On the one hand, I feel that an externally hosted blog is more useful from an availability perspective. It’s a pretty safe bet that the time you really want the information is when the machine running it is unavailable. Obviously if you choose this route make sure you use necessary security so you don't give anything away.

On the other hand, because we use WordPress we can have multiple web servers talking to the database and then use replication to create a hot-backup to another server. This works well on an internal basis as long as you have multiple machines to run MySQL, Apache and PHP on, and you don't have a weak point in your network.

The ideas are not new - afterall we’ve had monitoring systems for IT management for years – but the integration of the information into the mainstream of incoming information is. It takes a little longer to get the systems set up correctly to generate the information. I use Perl, PHP and MySQL for the majority of the code and monitoring system, which isn’t ideal if you don’t have the in-house experience to create the tools in the first place, but the basic mechanics shouldn’t be difficult for most IT departments to get their head around.

In the long run though, the RSS route makes monitoring and managing your network quite a lot easier. Even better, getting new members of staff to subscribe to it and getting individuals to monitor what’s going on has become easier.

Martin 'MC' Brown is a freelance writer and consultant. He can be contacted through his website MCslp.com