tracksolid_timescale_grafana_prod

kianiadee/tracksolid_timescale_grafana_prod

Fork 0

Commit graph

Author	SHA1	Message	Date
david kiania	e5b0e192d8	chore(repo): reorganize tree into migrations/ data/ legacy/ docs/ Group root-level files (accreted from incremental changes) by purpose without moving any deployment entrypoint or breaking imports: - migrations/ : numbered SQL 02-10 - data/ : source CSVs - legacy/ : superseded pre-_rev scripts + old pipeline notes (not deployed) - docs/{manuals,reference,reports}/ : loose manuals, references, reports - strip stray / * prefixes from 5 doc filenames - delete empty documents.txt / push_webhook.md Reference updates so nothing breaks: - run_migrations.py -> /app/migrations/<file> - run_migrations.sh -> $SCRIPT_DIR/migrations - import_drivers_csv.py -> data/<csv> - docker-compose.yaml -> runbook path comment - CLAUDE.md -> codebase map + inline doc references Deployed Python (3 services + ts_shared_rev + run_migrations) and the documented ops one-shots stay at root, preserving the flat-import layout and all documented commands. Verified: py_compile clean across all modules, every MIGRATIONS entry resolves under migrations/, CI-referenced paths (tests/, mypy targets, db_audit) and the grafana build context intact. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	2026-06-01 02:27:30 +03:00
David Kiania	e811dd8f34	feat(infra): add pgbouncer sidecar to cap tracksolid_db connections Some checks are pending Static Analysis / static (push) Waiting to run Details Tests / test (push) Waiting to run Details Phase 1 of the pgbouncer + pgAdmin rollout (runbook: 260507_pgbouncer_deployment.md). pgAdmin4 on the maintainer's laptop has been exhausting tracksolid_db's max_connections, cascading to pgcli and operations. Adds an internal-only pgbouncer service in transaction mode with a small backend pool (default 15) so admin-tool sprawl can no longer starve the ingest pipeline. No client cutover this round - ingest, Grafana, webhook, and backup all keep talking to timescale_db:5432 directly. SCRAM passthrough is wired via a new pgbouncer role + public.user_lookup() function (migration 10). The role is created with a placeholder password; sync_role_passwords() in run_migrations.py replaces it from PGBOUNCER_AUTH_PASSWORD on every container startup, mirroring the existing grafana_ro convention. Requires PGBOUNCER_AUTH_PASSWORD to be set in .env before deploy. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	2026-05-07 13:21:35 +03:00

Author

SHA1

Message

Date

david kiania

e5b0e192d8

chore(repo): reorganize tree into migrations/ data/ legacy/ docs/

Group root-level files (accreted from incremental changes) by purpose
without moving any deployment entrypoint or breaking imports:

- migrations/  : numbered SQL 02-10
- data/        : source CSVs
- legacy/      : superseded pre-_rev scripts + old pipeline notes (not deployed)
- docs/{manuals,reference,reports}/ : loose manuals, references, reports
- strip stray ** / *** prefixes from 5 doc filenames
- delete empty documents.txt / push_webhook.md

Reference updates so nothing breaks:
- run_migrations.py  -> /app/migrations/<file>
- run_migrations.sh  -> $SCRIPT_DIR/migrations
- import_drivers_csv.py -> data/<csv>
- docker-compose.yaml -> runbook path comment
- CLAUDE.md -> codebase map + inline doc references

Deployed Python (3 services + ts_shared_rev + run_migrations) and the
documented ops one-shots stay at root, preserving the flat-import layout
and all documented commands. Verified: py_compile clean across all modules,
every MIGRATIONS entry resolves under migrations/, CI-referenced paths
(tests/, mypy targets, db_audit) and the grafana build context intact.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

2026-06-01 02:27:30 +03:00

David Kiania

e811dd8f34

feat(infra): add pgbouncer sidecar to cap tracksolid_db connections

Static Analysis / static (push) Waiting to run

Details

Tests / test (push) Waiting to run

Details

Phase 1 of the pgbouncer + pgAdmin rollout (runbook:
260507_pgbouncer_deployment.md). pgAdmin4 on the maintainer's laptop has
been exhausting tracksolid_db's max_connections, cascading to pgcli and
operations. Adds an internal-only pgbouncer service in transaction mode
with a small backend pool (default 15) so admin-tool sprawl can no
longer starve the ingest pipeline.

No client cutover this round - ingest, Grafana, webhook, and backup all
keep talking to timescale_db:5432 directly. SCRAM passthrough is wired
via a new pgbouncer role + public.user_lookup() function (migration 10).
The role is created with a placeholder password; sync_role_passwords()
in run_migrations.py replaces it from PGBOUNCER_AUTH_PASSWORD on every
container startup, mirroring the existing grafana_ro convention.

Requires PGBOUNCER_AUTH_PASSWORD to be set in .env before deploy.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

2026-05-07 13:21:35 +03:00

2 commits