Sovran_Systems/Sovran_SystemsOS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 11 Wiki Activity

nextcloud, postgresql: fix Nextcloud 35 DB warnings on 32 GB hosts

Browse Source 
Nextcloud 35's Database checks flag three Performance issues out of the
box: buffer cache hit ratio ~96% (wants 99%+), 100k+ dead tuples, and
million-plus sequential scans on oc_mail_tags / oc_guests_users.

Root causes in Sovran: stock 128MB shared_buffers, stock 60s autovacuum
naptime, APCu file locking, and db:add-missing-indices running exactly
once at install time (never on upgrades or app installs).

Size Postgres for the README's Server + Desktop recommendation (32 GB
RAM, NVMe): 2GB shared_buffers, 12GB effective_cache_size, 512MB
maintenance_work_mem, 32MB work_mem, 4GB max_wal_size, 30s autovacuum
naptime with 4 workers. shared_buffers stays below the 25% rule because
Postgres shares the box with bitcoind, Electrs, LND, MariaDB and PHP-FPM.

Scope the aggressive autovacuum to nextclouddb via ALTER DATABASE so the
shared matrix-synapse DB keeps the milder cluster defaults.

Add a local Redis (127.0.0.1:6379, Nextcloud only) and move
memcache.distributed/locking to Redis; migrate existing installs with a
one-shot since nextcloud-init never re-runs.

Add a weekly nextcloud-db-maintenance timer (VACUUM ANALYZE +
db:add-missing-*) so upgrades and later app installs can't regress the
checks again.

Note: shared_buffers needs one 'systemctl restart postgresql', which
briefly takes down both Nextcloud and Matrix. Everything else is
reload-only or scoped to nextclouddb.
This commit is contained in:
Sovran Contributor
2026-09-21 13:59:52 -05:00
committed by naturallaw777
parent 3341659a0c
commit 09d4cc9b83
2 changed files with 148 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
configuration.nix
+36
View File
@@ -165,6 +165,14 @@
programs.fish = { enable = true; promptInit = "fastfetch"; };
# ── PostgreSQL base ────────────────────────────────────────
# Shared cluster for Nextcloud (nextclouddb) + Matrix Synapse.
# Sized for the README's Server + Desktop recommendation (32 GB RAM,
# 500 GB NVMe OS + 2 TB NVMe timechain). Postgres shares the box with
# Bitcoin Core, Electrs, LND, MariaDB, PHP-FPM and GNOME, so
# shared_buffers stays below the 25%-of-RAM dedicated-server rule.
# Fixes Nextcloud 35 Database checks (pg.cache_hit_ratio,
# pg.dead_tuples). Override in custom.nix for other hosts, e.g.:
# services.postgresql.settings.shared_buffers = lib.mkForce "512MB";
services.postgresql = {
enable = true;
authentication = lib.mkForce ''
@@ -172,6 +180,34 @@
host all all 127.0.0.1/32 trust
host all all ::1/128 trust
'';
settings = {
# Memory — fixes low buffer cache hit ratio (stock default is
# 128MB shared_buffers). effective_cache_size is only a planner
# hint, not an allocation, so it can be generous.
# NOTE: changing shared_buffers requires a Postgres restart.
shared_buffers = "2GB";
effective_cache_size = "12GB";
maintenance_work_mem = "512MB";
work_mem = "32MB";
wal_buffers = "64MB";
# Checkpoints — spread write bursts out on NVMe. Reload-only.
min_wal_size = "1GB";
max_wal_size = "4GB";
checkpoint_completion_target = 0.9;
# Autovacuum — the stock 60s naptime can't keep up with
# Nextcloud's and Synapse's write-heavy tables (filecache,
# activity, jobs, state). Reload-only.
autovacuum_naptime = "30s";
autovacuum_vacuum_scale_factor = 0.05;
autovacuum_analyze_scale_factor = 0.025;
autovacuum_max_workers = 4;
# NVMe planner assumptions (README: NVMe OS + data disks).
random_page_cost = "1.1";
effective_io_concurrency = 200;
};
};
# ── Backups ────────────────────────────────────────────────