mirrors/mdlayher-homelab
1
0
Fork 0
mirror of https://github.com/mdlayher/homelab.git synced 2024-12-14 11:47:32 +00:00
Code Activity

nixos/servnerr-3: switch to zrepl replication

Browse source 
Signed-off-by: Matt Layher <mdlayher@gmail.com>
This commit is contained in:
Matt Layher 2022-02-14 16:17:52 -05:00
parent 4cc210085b
commit 9d2c43eb6c
No known key found for this signature in database
GPG key ID: 77BFE531397EDE94
2 changed files with 38 additions and 66 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
nixos/servnerr-3/prometheus.nix
View file

@ -145,7 +145,7 @@ in {
(blackboxScrape "http_2xx" "15s" [
"http://living-room-myq-hub.iot.ipv4"
"https://grafana.servnerr.com"
])
])
# Netlify can occasionally be flappy, so check it less often.
(blackboxScrapeJobName "http_2xx_netlify" "http_2xx" "1m" [
"https://corerad.net/status"

102
nixos/servnerr-3/storage.nix
View file

@ -16,20 +16,29 @@ let
client_identity = "local";
};
# Let zfs-auto-snapshot manage the snapshotting.
snapshotting.type = "manual";
# Snapshot every 15 minutes.
snapshotting = {
type = "periodic";
prefix = "zrepl_";
interval = "15m";
};
pruning = {
# Keep all primary snapshots, zfs-auto-snapshot manages them.
keep_sender = [{
type = "regex";
regex = ".*";
}];
# Keep the last few snapshots for each dataset for disaster recovery.
keep_receiver = [{
type = "last_n";
# 8 snapshots for each dataset.
count = 8;
}];
keep_sender = [
# Keep snapshots that are not already replicated.
{
type = "not_replicated";
}
# Keep manual snapshots.
{
type = "regex";
regex = "^manual_.*";
}
# Keep time-based bucketed snapshots.
keepGrid
];
# Keep the same automatic snapshots as source.
keep_receiver = [ keepGrid ];
};
});
@ -79,6 +88,21 @@ let
# Generate the zrepl sink job name for a target zpool.
sinkName = (zpool: "sink_${zpool}");
# Keep time-based bucketed snapshots.
keepGrid = {
type = "grid";
# Keep:
# - every snapshot from the last hour
# - every hour from the last 24 hours
# - every day from the last 2 weeks
# - every week from the last 2 months
# - every month from the last 2 years
#
# TODO(mdlayher): verify retention after a couple weeks!
grid = "1x1h(keep=all) | 24x1h | 14x1d | 8x7d | 24x30d";
regex = "^zrepl_.*";
};
in {
# ZFS filesystem mounts.
#
@ -150,31 +174,6 @@ in {
# Scrub all pools regularly.
autoScrub.enable = true;
# Roll up snapshots for long periods of time, we have storage to burn.
autoSnapshot = {
enable = true;
# Debug output, keep zero-sized snapshots, parallel snapshots, UTC
# timestamp, verbose logging. Only snapshot primary.
flags = "-d -k -p -u -v -P primary";
# High frequency snapshots. For quickly rolling back unintended changes,
# so we don't keep very many.
#
# No frequent snapshots; unneeded.
frequent = 0;
# Every hour for 4 hours.
hourly = 4;
# Beyond this point, retain more snapshots for long-term archival.
#
# Every day for 2 weeks.
daily = 14;
# Every week for 2 months.
weekly = 8;
# Every month for 2 years.
monthly = 24;
};
# ZED configuration.
zed = {
enableMail = false;
@ -221,31 +220,4 @@ in {
};
};
};
# Manual systemd unit and timer to trigger zrepl jobs. We use
# zfs-auto-snapshot integrated into NixOS instead of zrepl's built-in
# automatic snapshotting, so we have to signal replication manually.
#
# TODO(mdlayher): push upstream into services.zrepl configuration.
systemd = {
services.zrepl-signal-jobs = {
serviceConfig.Type = "oneshot";
path = with pkgs; [ zrepl ];
script = ''
zrepl signal wakeup ${pushName "secondary"}
# zrepl signal wakeup ${pushName "backup0"}
# zrepl signal wakeup ${pushName "backup1"}
'';
};
timers.zrepl-signal-jobs = {
wantedBy = [ "timers.target" ];
partOf = [ "zrepl-signal-jobs.service" ];
timerConfig = {
# TODO(mdlayher): this is a hack; try to offset zrepl jobs from
# zfs.autoSnapshot running every hour.
OnCalendar = "*:20";
Unit = "zrepl-signal-jobs.service";
};
};
};
}