3 ms·
Would like more info on [1]. I work in the area of cutting costs (also privately - gave up coffee as it was a waste of money, for me): with money being finite,
by InCityDreams 3y ago
Would like more info on [1]. I work in the area of cutting costs (also privately - gave up coffee as it was a waste of money, for me): with money being finite, what you can save is often worth more than what you can earn.
- ChuckMcM 3y agoIt wasn't particularly transferable outside of Google's scale. They have continual failures in their fleet of machines which cycle through a repair queue. It's fairly automated, dispatching bulk repair requests to people in the data centers. I managed to get some process changes and code changes through that allowed machines with alleged disk errors to be processed while still in production and have the resources returned after the disk was verified good and reformatted. When you calculated depreciation costs for the equipment that had been waiting for repairs before, and now never left production, it was well over $10M/year of depreciation cost. (there were other cost savings as well like the over provisioning needs of a cluster were lower , etc etc) They weren't wrong in that it wasn't particularly technically challenging, but getting a change like that through the organization had been tried and failed several times before I was given that task and succeeded. That skill, being an effective change agent in a large organization, was not even on their radar of useful skills at that time.