Sourcegraph logs warnings about this on startup otherwise. Unclear
to what degree it really affects operation though.
Change-Id: I6ee7c5358631aafd9a7f8155150361bf7499314d
Reviewed-on: https://cl.tvl.fyi/c/depot/+/3098
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
.. this is actually likely not disabling it for some pages, that will
need this to be copy & pasted, but it's hard to tell just from the
nginx docs. We'll make sure after deploying.
Change-Id: I2fa6e31ca10835a206673b858594fa071e729d82
Reviewed-on: https://cl.tvl.fyi/c/depot/+/3020
Tested-by: BuildkiteCI
Reviewed-by: lukegb <lukegb@tvl.fyi>
It needs to refer to this by full path of course.
Change-Id: I911c876ba18877681accb722426314d92b9f2318
Reviewed-on: https://cl.tvl.fyi/c/depot/+/3042
Reviewed-by: lukegb <lukegb@tvl.fyi>
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
This will require the daemon to be running when launching GC, but
won't start it if it happens to not be running for some reason.
Change-Id: If48fe336030173f028428fc00a81d339ef4b8bce
Reviewed-on: https://cl.tvl.fyi/c/depot/+/3015
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
Adds a module that automatically collects garbage based on disk space
thresholds, and configures it to run hourly on whitby.
This is implemented as an alternative to cl/2937, which I've been told
uses a Nix feature that doesn't actually work.
Under-the-hood this is simply a systemd timer running a shell script
which checks available disk space and runs GC when necessary.
Change-Id: I3c6b5de85b74ea52e7e16c53f2f900e0911c9805
Reviewed-on: https://cl.tvl.fyi/c/depot/+/3014
Tested-by: BuildkiteCI
Reviewed-by: lukegb <lukegb@tvl.fyi>
Enables a Grafana service pointing to whitby's local Prometheus
instance, accessible at status.tvl.su.
I've no idea how to configure Grafana and if it's possible to link it
to CAS, but we'll see about that later.
Notes:
* the explicit fixpoint for whitby config has been removed as we
have the `config` parameter available now
* backups are enabled for the Grafana storage location
Change-Id: If5ffe0c1a3378d1c88529129487c643642705fd2
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2948
Tested-by: BuildkiteCI
Reviewed-by: grfn <grfn@gws.fyi>
Instead of having two ways of accessing the path to the depot (one of
which was stuttering, depot.depotPath) we settle on only one:
depot.path.
This was mostly used for NixOS module imports.
Co-Authored-By: Florian Klink <flokli@flokli.de>
Change-Id: I2c0db23383fc34f6ca76baaad4cc4af2d9dfae15
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2962
Tested-by: BuildkiteCI
Reviewed-by: grfn <grfn@gws.fyi>
Reviewed-by: sterni <sternenseemann@systemli.org>
Rename my //users directory and all places that refer to glittershark to
grfn, including nix references and documentation.
This may require some extra attention inside of gerrit's database after
it lands to allow me to actually push things.
Change-Id: I4728b7ec2c60024392c1c1fa6e0d4a59b3e266fa
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2933
Tested-by: BuildkiteCI
Reviewed-by: tazjin <mail@tazj.in>
Reviewed-by: lukegb <lukegb@tvl.fyi>
Reviewed-by: glittershark <grfn@gws.fyi>
Splits //ops/nixos into:
* //ops/nixos.nix - utility functions for building systems
* //ops/machines - shared machine definitions (read by readTree)
* //ops/modules - shared NixOS modules (skipped by readTree)
This simplifies working with the configuration fixpoint in whitby, and
is overall a bit more in line with how NixOS systems in user folders
currently work.
Change-Id: I1322ec5cc76c0207c099c05d44828a3df0b3ffc1
Reviewed-on: https://cl.tvl.fyi/c/depot/+/2931
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
Reviewed-by: glittershark <grfn@gws.fyi>
