Thursday, 23 January 2014

2015...scratch that...2014 project update

After finding out that the deadline for the project to install wireless across the whole campus has been brought forward by a year my boss has told me to only concentrate on the wireless network, in particular, the rollout project.

So since I've been back from Christmas most of my time has been taken up by project planning, meetings and wireless surveys.

For surveys I have been using AirMagnet Survey (with Planner also installed), which is an excellent tool.


For those who don't know AirMagnet was acquired by Fluke Networks and provides the ability to perform passive surveys (get a picture of all RF), active surveys (connect to AP or SSID), or iPerf surveys (for testing throughput). Our version also incorporates Planner that can be used to do predictive modelling. If you've got money to burn you can also add spectrum analysis into the package.

No matter how good the tool though, the more accurate the survey the longer it takes, and currently I have no idea how I'm going to get all the validation surveys done from the previous installs, let alone survey the whole of campus before the next installs should commence.

Currently, access point wise, I am looking at the Cisco 3602 as this offers beamforming for 802.11n clients and uses four transmitters (Our previous AP's were the 1142 and 3502 that has ClientLink 1.0 - 802.11a/g with 2 transmitters). It also works with our current version of wireless software (7.2). We are hoping to upgrade to 7.6 at some point soon due to the massive benefits including 802.11ac support and AVC. Then we will be able to look at the 3700, Cisco's 802.11ac access point (a module can be purchased for the 3600 series to add 802.11ac but it's not the ideal solution).

Also, when deploying all these access points I need the time to check them, asset mark them, put them in a hardware database, record MAC addresses, serial numbers, order numbers and so on. All the boring stuff.

As our helpdesk software doesn't really do what we want it to do from a project management point of view (and we don't have the time to wait for it to be incorporated and tested) we have started using Trello to plan and organise the project.

So far this has worked great. I won't go into all the features, as there are many, but I have provided the link above if you are interested and what really pleases the bosses is that it is free.

The advantages for us are that it is really easy for everyone to collaborate on the board, there is an endless activity log and archive should things go wrong, it works great with attachments and there are iOS and android apps for using it on the move (it's for a wireless project, we will probably need the info most when on the move).

There are also colour coded labels, the ability to add checklists and due dates/calendar options. You can also assign to users like you would through a helpdesk and you get notified if somebody mentions you on the board.

The picture will be difficult to see but the way we have decided to use it is to have a range of tables or columns (called lists in Trello) that have all the stages of the project. Then each building is entered on what Trello calls a card. Surveys, pictures, checklists, notes and more can be added to these cards. Then we move the card along depending on what stage we are at until the building is complete. The lists are:

  • Buildings yet to start
  • Defining/designing
  • Surveying
  • Installing
  • In Service
  • Post-install survey
  • Done
Then we have separate board for teaching spaces where we require per-seat coverage (enough throughput to allow them to do what they want to use wireless for). As well as the above there are lists labelled:
  • No coverage
  • Simple coverage
  • Per seat coverage

As I say it's working really well and if we get external people to help with the project (contractors, students, etc) we can invite them to the boards as the accounts are free.

I may write more about Trello in a future post.

So that's pretty much where I'm at.

On a side note, we are about to lose our redundancy for wireless controllers.

Currently we have a controller for campus (with a back up in a different data centre) and one for the student villages (also with a backup on a different site). With the installs currently going in we are going to reach the WiSM2 limit of 1000 AP's on the main campus controller and have to start using the backup. They are too expensive to get a further two controllers in case a router or fibre link goes down so we'll have to live with that for now.

No comments: