ignoring messages that a database server is nearly out of space. and loosing nearly all database servers after ignoring the message and then telling that the database is bad and they switched to mysql? well it wasn't just that, they also did wierd things with a relation database and they still do with mysql. I'm not sure but actually keeping the database servers happy should be the no 1 job of an SRE team, since the data is the most important thing to any company.

Asking why Twitter is so bloated gets 300 upvotes, but asking why Uber SRE is so bloated gets downvoted? My, what a fickle crowd HN is.

My assumption was that Uber SRE's played roles similar to Google SRE's. If that's not the case, someone can please say so instead of just downvoting? If you lump all ops people (DBA, sysadmin, network, storage) into the SRE term then 150 isn't huge. But if it's mainly sysadmin and devops, that's really high for a company that's mainly in the cloud.