Live data from Hacker News

Monarch: Google’s Planet-Scale In-Memory Time Series Database

micahlerner.com

131–133 of 133 posts

Re: Monarch: Google’s Planet-Scale In-Memory Time Series Database

#131

Earlier quoted context omitted.

Collect directly from shard-level prometheus then aggregate using /federate at another level. That’s how thanos also works afaik

This still requires the query to fit on a single node at some point unless you're doing multi-level aggregation. Monarch does that. Being able to push expensive, highly-parallelizable aspects of the query down into the leaves is a massive benefit.

Yep, in theory it’s great if you can afford a team of highly qualified engs to run it like google and dealing with huge metrics. In practice I found it was not hard to crash a monarch mixer (at least some years ago) + most orgs won’t be able to afford nor take advantage of it

Re: Monarch: Google’s Planet-Scale In-Memory Time Series Database

#133
post #3

I broke this once several years ago. I even use the incident number in my random usernames to see if a Googler recognizes it.

Could someone elaborate on this

https://status.cloud.google.com/incident/google-stackdriver/... was one of the public incident reports around it.
Post reply on HN