|
This is true, but I am unsure of the fix prescribed here. For me, I had experience with a project like this - a large, extremely bureaucratic company tasked my team with a nearly impossible task, I think in an effort to lay off the team - which eventually happened. I fought with tooth and nails and was already doing fairly well in my role, so they gave me a shot and assigned the task, which I won't belabor the details of, but it was basically to implement a fundamental service that every application in the company would use to authenticate to the backend system (plus a lot of other reliability/availability guarantees). The problem was, with their architecture being a central hub cluster of servers that provided core services to hundreds of edge servers around the world, it would have been fine to just throw this service into the central server and call it a day. However, the eggheads in executive leadership felt this was not acceptable, so the requirement was to make a new "central" cluster to connect to not only all the edge clusters in the company, but all the backend office/admin stuff as well. Problem was - the people who wired up the networking and everything else with the central server had been gone for 10+ years. Barely anyone even knew how it worked, and unfortunately as I took this on, the network team got laid off, replaced by people who also had no clue how it worked. As we connected more and more services to this new hub, a bunch of skeletons emerged - the funniest being a server cluster in a region and account that no one had access to, or knew what it did, other than when it crashed other servers had issues. No one had accessed it for years. That was fun hacking into. There were tons of hurdles like this at every step, basically being that this touched so many teams in so many different areas of the infrastructure, the organizational hurdles trying to even get the information you need to do the job required a ridiculous amount of heroics. I hated it. Basically, you need to create urgency any way you can - whether this is by breaking things, horse trading, political maneuvers, begging, intimidation - your livelihood is on the line. Other teams sensed what a difficult ask this was so early going was extremely difficult getting cooperation from the 30+ teams this touched. Anyway at the end of the day I was able to do it by finally escalating urgency to the executive level and they made it a priority one quarter and it quickly got finished. If they hadn't, it would have been very difficult. They had the issue of having too many layers of management below them and had no idea, all they heard is "why is this project you said would take 2 months taking almost 2 years" and a bunch of manager speak trying to explain why. I would never, ever want to work on any project like that again. 2 years to complete, should have taken 2 months if working as an IC, 2 weeks with a full competent team and good management. There were some good things that came out of it though - processes got improved, collaboration improved, and we were able to use it as a chance to refactor the IAC in a way you could deploy these hub servers again in a much easier way that didn't require the ridiculous amount of detective work to figure out the first time. Oh yea, forgot to mention the actual application took about 2 days to configure. All the rest of it was what took 2 years. |