[Openinfralabs] Prometheus long-term storage

[Lars Kellogg-Stedman]

Hey,

We're currently evaluating Prometheus at the MOC right now, and one of
our open questions is what platform we should be using for providing
(a) a global view of multiple Prometheus instances (b) some form of HA
(basically, we want to be able to take a prometheus host out of
service for maintenance without losing metrics) and (c) long term
storage of metrics.

I've looked at both Thanos and Victoria Metrics, and from both a
design and performance perspective Victoria Metrics seems better...but
Thanos seems to have more mindshare among people I've spoken with, and
looking online for solid comparisons between the two didn't turn up
much except some finger-pointing from developers of both solutions
(and I haven't even tried to evaluate other options yet, like Cortex,
M3DB, etc).

Are there folks on this list now who have opinions about these
options?

-- 
Lars Kellogg-Stedman <lars at redhat.com> | larsks @ {irc,twitter,github}
http://blog.oddbit.com/                | N1LKS