Managing IT projects in spreadsheets is painful: part 1

  by Simon Woodward | 05 May 2009 | Permalink |

The information that Tideway gathers about the IT estate is very well suited to supporting IT projects. The hardware and software inventory details provide valuable information to support designing and capacity planning of new environments (virtual environments or new datacentres), while the dependency information makes planning the project a lot simpler. Deciding which hosts to change/move together, and in which order to process the changes is a lot simpler when you know which business services are provided by what infrastructure.

The work that has gone into the next release of Tideway (currently codenamed Titan) has been in support of this use case. In order to make the product even more useful for doing an IT baseline, including inventory and dependency data, we’ve made Tideway:

• easy to roll out into a new estate, or into a new portion of the estate (here)
• provide immediate, intuitive information to help baseliners understand the estate (here).
• support the process of obtaining knowledge from the company’s IT staff (here).

These pieces of functionality were discussed by me in previous blog posts (here and here) as they were completed, and I’m happy to say that this post is similar. We’ve just finished the next bit of work, which enables the baseliner to divide the estate up how they wish, capturing their discovered knowledge without having to export data to a spreadsheet for further processing.

Spreadsheets, spreadsheets, everywhere…

Everybody that I speak in an IT department, IT project management or even our consultants who deal with such people complain about the ever-present spreadsheet. These sheets are generated by munging together data from manual audits, exports from automatic discovery tools and possibly other sources. Once generated, they’re out of date within hours as the estate continues to evolve. Using them for planning the project is awkward because they’re divorced from their source data – going back and checking anything, or finding extra data involves laborious cross-referencing to look items up in the source data. Collaborating on them is hard – the team members make changes to private copies, and then have to meet to merge their changes back into the master. This is the pain that we’re trying to alleviate by providing the required management functionality directly in Foundation, so that exports to spreadsheets are either not necessary, or are only necessary much later.

People have three things that they use spreadsheets in these projects for:

1. They want to group the hosts in the spreadsheet. The groups are either manual (“this group contains this one, this one and… um….that one”) or dynamic (“this group contains all of the hosts in my Paris subnet (ie. anything with an address in the range 172.17.0.0/16)”).
2. They want somewhere to capture additional knowledge that they’ve gathered about the group or about individual hosts. They can add columns, write notes and do free text searches on the captured data.
3. They want a way to hand the completed baseline off to the team that will execute the project plan. The spreadsheet forms part of the final documentation set.

A related need is that of munging together data from different sources – pulling your manual audit data into the spreadsheet, for example. This is already possible via our new CSV import page: see here for details.

We’re planning on providing all of the above functionality directly in Tideway. Our first bit of work towards this is completed; users are now able to:

• Create groups of hosts (or other items)
• Add items to and remove items from that group (either via a report or from a visualisation)
• Establish a working set of groups (mark a number of groups as “favourite”)
• See items’ group memberships from reports, making grouping of them easier
• Export groups of items

Together, this fulfils the first use-case above: manual grouping of items. We do intend to do dynamic grouping before we release Titan, and I still hold out hope for the ability to capture manual data!

This functionality is now available in pre-release 3. See here for documentation about these features, and please post feedback on our forums!