Received: from malur.postgresql.org ([217.196.149.56]) by arkaria.postgresql.org with esmtps (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.98.2) (envelope-from ) id 1xQGaA-00000001HaJ-25T0 for pgsql-announce@arkaria.postgresql.org; Mon, 09 Nov 2026 04:02:12 +0000 Received: from localhost ([127.0.0.1]:46760 helo=malur.postgresql.org) by malur.postgresql.org with esmtp (Exim 4.98.2) (envelope-from ) id 1xQGa9-00000007hFf-3ks5 for pgsql-announce@arkaria.postgresql.org; Mon, 09 Nov 2026 04:02:09 +0000 Received: from magus.postgresql.org ([2a02:c0:301:0:ffff::29]:57960) by malur.postgresql.org with esmtps (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.98.2) (envelope-from ) id 1xQGa9-00000007hFD-0WJ9 for pgsql-announce@lists.postgresql.org; Mon, 09 Nov 2026 04:02:09 +0000 Received: from mahout.postgresql.org ([2001:4800:3e1:1::227]) by magus.postgresql.org with esmtps (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.98.2) (envelope-from ) id 1xQGa6-00000001978-16GB for pgsql-announce@lists.postgresql.org; Mon, 09 Nov 2026 04:02:08 +0000 DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=postgresql.org; s=20171124; h=Message-ID:Date:Reply-To:From:To:Subject: MIME-Version:Content-Type:Sender:Cc:Content-Transfer-Encoding:Content-ID: Content-Description:In-Reply-To:References; bh=1gGuKrNFdWhCoMcIlSRom4UiMAOtEJbRO7giIEHwI4o=; b=i4YA4gRQg1H8EKfb6yZR4vKr89 7/aALGtme6jUe+AO9+rRIwUs7nmFIJT9XSSid3KY8IdvC1JcWrXs9Y5saIfCbcKmLwD5ShySXvo9R WFFhYwF3l94ckxGWS4pFR/ByRDCHUZbycVl/FzAqjOwb+XoRTXs/eZKv2kgKnrZxwwMVBxASAOO3I Z3ugU3tfz2ru6DTDJJLehxZVUSsm/OhMpG1Gf6XtyR/DuWMApbYU0gOxYJuAi+CkVnpRdjDJQ9L6G sLDKyeeWyNur8m8quwZSu6lTWH1NTMa/xm7QfOUp/WMP2sX07gIpJ7V6RBjo6/HEOGbw0cc3x8Uo4 1BeZAI7g==; Received: from wrigleys.postgresql.org ([2a02:16a8:dc51::60]) by mahout.postgresql.org with esmtps (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.98.2) (envelope-from ) id 1xQGa4-00000002kTv-3eNx for pgsql-announce@lists.postgresql.org; Mon, 09 Nov 2026 04:02:05 +0000 Received: from localhost ([127.0.0.1] helo=wrigleys.postgresql.org) by wrigleys.postgresql.org with esmtp (Exim 4.98.2) (envelope-from ) id 1xQGa3-00000006gpA-0N38 for pgsql-announce@lists.postgresql.org; Mon, 09 Nov 2026 04:02:03 +0000 Content-Type: multipart/alternative; boundary="===============7676022457510646191==" MIME-Version: 1.0 Subject: Dasha 1.10: Patroni auto-discovery, wait analysis and replication slot forecasts for PostgreSQL To: PostgreSQL Announce From: Dasha via PostgreSQL Announce Reply-To: d.bulashev@gmail.com Date: Mon, 09 Nov 2026 04:01:43 +0000 Message-ID: <179419690348.595140.7659248240103249087@wrigleys.postgresql.org> X-Auto-Response-Suppress: All Auto-Submitted: auto-generated X-pglister-tags: related X-pglister-tagsig: a7eb01688c6c6f9f6244ef173754e7b0991e0d6e493648a1eafa25b1cf9ba2fc List-Id: List-Help: List-Subscribe: List-Post: List-Owner: List-Archive: Archived-At: Precedence: bulk --===============7676022457510646191== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Dasha is an open source performance dashboard for PostgreSQL fleets. It con= nects to your clusters with a read-only role, shows what the databases are = doing right now, and explains what to do about it. Nothing is installed on = the database hosts: no agent, and no extension beyond the ones you already = have. One instance serves many clusters and all of their replicas. Most PostgreSQ= L statistics are per-instance and are not replicated, so Dasha reads every = host of a cluster and reasons over the combined picture. An index that look= s unused on the primary may be serving the entire read workload on a standb= y. ### New since the last announcement * **Patroni clusters from etcd v3 or Consul.** Dasha reads the keys Patroni= keeps in its DCS and turns every Patroni scope into a cluster: the members= become hosts and the databases are listed from PostgreSQL. New clusters, l= ost replicas, created and dropped databases show up within the refresh inte= rval. The DCS is only read, never written. User, password, port and databas= e filters can be overridden per scope, and TLS with client certificates is = supported for both etcd and Consul. Member names appear in the host selecto= r next to the address. * **Standby cluster links.** A Patroni standby cluster shows the cluster it= replicates from, and the primary shows its standbys, on the Summary and Re= plication pages. Clusters are matched by the PostgreSQL system identifier, = so a standby behind a VIP is found too. * **Activity & Waits** (PostgreSQL 14+). A live chart of active sessions by= class: CPU, wait event type (`IO`, `Lock`, `LWLock`, ...) and idle in tran= saction, sampled from `pg_stat_activity` every 1-10 seconds. Below it are t= he top statements, sessions, databases, users, applications and wait events= ; a click on a value adds a filter, and a statement links to its query repo= rt and the lock tree. Statements are grouped by `query_id`. No extension is= needed, and each Dasha instance reads a host's `pg_stat_activity` at most = once per second however many people watch the page. * **Replication slot invalidation forecast.** Dasha measures the WAL rate o= ver a chosen window (30 seconds to 5 minutes) and shows, for every slot, th= e WAL it retains, the remaining `max_slot_wal_keep_size` reserve and the ti= me until the slot is lost. The measurement takes two LSN samples by the ser= ver clock and holds no connection in between. * **COPY progress** (PostgreSQL 14+). Table, source type, bytes and rows pr= ocessed, rows excluded and skipped, and a percentage with ETA when the tota= l is known: the file size for `COPY FROM` a server-side file, the row estim= ate for `COPY TO`. * **Health Score from the stock exporters.** The time-series mode of the He= alth Score now works with plain `postgres_exporter`, `node_exporter` and `p= gbouncer_exporter`, so an existing Prometheus or VictoriaMetrics setup is e= nough. Metrics of Patroni clusters are mapped automatically by scope and me= mber name. * **Cold tables are not penalized.** Tables with no writes for longer than = an inactivity threshold (7 days by default) no longer lower the score for d= ead rows and missing vacuum. A separate recommendation lists those that nee= d a single `VACUUM (FREEZE, ANALYZE)`, and the Maintenance page can filter = all, active or cold tables. ### Features * **Health Score:** a composite 0-100 score per instance across eight categ= ories, with prioritized recommendations tied to a specific database. Option= ally computed from Prometheus/VictoriaMetrics time series, with a trend and= a seasonal baseline. * **Query analysis:** top queries by execution time and WAL, running and bl= ocked queries with wait events, and `pg_stat_statements` snapshots. Snapsho= ts can be saved to a storage database, shared by URL, and diffed against an= other snapshot or against live data. * **Index advisor and analysis:** missing B-tree indexes proposed from `pg_= stat_statements` of every host, optionally backed by `auto_explain` plans f= rom the log. Bloat, duplicates, invalid indexes, and an answer to "may I dr= op this index?" that accounts for statistics resets, per-host counters and = partitioned indexes rather than trusting `idx_scan =3D 0`. * **Schema checks:** 17 checks for structural defects, such as sequences cl= ose to exhaustion, tables without a primary key or a `public` schema anyone= can create objects in, each with a severity and remediation SQL. * **I/O analysis** built on `pg_stat_io` (PostgreSQL 16+), with WAL on Post= greSQL 18. * **Log insights:** log search in OpenSearch, Elasticsearch, VictoriaLogs a= nd Yandex Managed PostgreSQL, event categories, `auto_explain` plan trees w= ith findings, and plan regressions between two time windows. * **Automatic snapshots:** a separate daemon captures query and instance-wi= de lock-contention snapshots on activity spikes or primary/replica role cha= nges. The evidence is there the next morning when someone asks what happene= d at 03:00. * **Tables and maintenance:** size with a TOAST breakdown, per-table descri= be, and autovacuum and wraparound thresholds computed with PostgreSQL's own= formula. Also progress of VACUUM, ANALYZE, CREATE INDEX, CLUSTER, base bac= kups and COPY, and hot tables and indexes derived from scheduled delta snap= shots. * **Cluster discovery:** a static list, Patroni clusters from etcd or Consu= l, or Yandex Managed Service for PostgreSQL; databases are discovered from = the cluster itself. * **Access control:** three auth modes (open, static API keys, or OIDC with= encrypted session cookies), Casbin RBAC, personal access tokens and per-id= entity rate limiting. * **MCP connector:** a read-only MCP server that gives AI assistants the fl= eet diagnostics. It has 35 tools, including index recommendations, schema c= hecks, I/O, log search, plan regressions and live activity sampling. It als= o has 5 prompts, per-user token passthrough and an embedded knowledge base = of the scoring rules. The interface is available in English, Russian and German. ### Requirements * PostgreSQL 14 or newer (tested on 14-19b) for the monitored clusters, rea= chable with a role holding `pg_monitor` and `CONNECT` on every monitored da= tabase. * `pg_stat_statements` (or `pgpro_stats` on Postgres Pro) for query statist= ics; `pgstattuple` is optional. * Activity & Waits needs `compute_query_id` set to `on`, or to `auto` with = `pg_stat_statements` or `pgpro_stats` preloaded. * The I/O page requires PostgreSQL 16 or newer. * Optional: etcd v3 or Consul read access for Patroni discovery. * Optional: a log source (OpenSearch, Elasticsearch, VictoriaLogs or Yandex= Managed PostgreSQL) for log search and plan insights, with `auto_explain` = enabled for plan analysis. * Optional: a separate PostgreSQL database for snapshots, hot-object histor= y and personal access tokens. Everything else works without it. ### Documentation * [README](https://github.com/dbulashev/dasha#readme): configuration, monit= oring role privileges, deployment, Helm values * [Patroni discovery](https://github.com/dbulashev/dasha/blob/main/doc/en/c= onfiguration.md#patroni-clusters-from-etcd-or-consul-optional) * [Health Score scoring model](https://github.com/dbulashev/dasha/blob/main= /README-health-score.md) * [Changelog](https://github.com/dbulashev/dasha/blob/main/CHANGELOG.md) ### License Dasha is licensed under the GNU General Public License v3.0; see [LICENSE](= https://github.com/dbulashev/dasha/blob/main/LICENSE). Project page: [github.com/dbulashev/dasha](https://github.com/dbulashev/das= ha) --===============7676022457510646191== Content-Type: text/html; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Dasha 1.10: Patroni auto-discovery, wait analysis and replicatio= n slot forecasts for PostgreSQL
 

Dasha 1.10: Patroni auto-discovery, wait analysis and replication slo= t forecasts for PostgreSQL

Dasha is an open source performance dashboa= rd for PostgreSQL fleets. It connects to your clusters with a read-only rol= e, shows what the databases are doing right now, and explains what to do ab= out it. Nothing is installed on the database hosts: no agent, and no extens= ion beyond the ones you already have.

One instance serves many clusters and all o= f their replicas. Most PostgreSQL statistics are per-instance and are not r= eplicated, so Dasha reads every host of a cluster and reasons over the comb= ined picture. An index that looks unused on the primary may be serving the = entire read workload on a standby.

New since the last announcement=

  • Patroni= clusters from etcd v3 or Consul. Dasha reads the keys Patroni kee= ps in its DCS and turns every Patroni scope into a cluster: the members bec= ome hosts and the databases are listed from PostgreSQL. New clusters, lost = replicas, created and dropped databases show up within the refresh interval= . The DCS is only read, never written. User, password, port and database fi= lters can be overridden per scope, and TLS with client certificates is supp= orted for both etcd and Consul. Member names appear in the host selector ne= xt to the address.
  • Standby= cluster links. A Patroni standby cluster shows the cluster it rep= licates from, and the primary shows its standbys, on the Summary and Replic= ation pages. Clusters are matched by the PostgreSQL system identifier, so a= standby behind a VIP is found too.
  • Activit= y & Waits (PostgreSQL 14+). A live chart of active sessions by= class: CPU, wait event type (IO, Lock, LWL= ock, ...) and idle in transaction, sampled from pg_stat_activi= ty every 1-10 seconds. Below it are the top statements, sessions, da= tabases, users, applications and wait events; a click on a value adds a fil= ter, and a statement links to its query report and the lock tree. Statement= s are grouped by query_id. No extension is needed, and each Da= sha instance reads a host's pg_stat_activity at most once per = second however many people watch the page.
  • Replica= tion slot invalidation forecast. Dasha measures the WAL rate over = a chosen window (30 seconds to 5 minutes) and shows, for every slot, the WA= L it retains, the remaining max_slot_wal_keep_size reserve and= the time until the slot is lost. The measurement takes two LSN samples by = the server clock and holds no connection in between.
  • COPY pr= ogress (PostgreSQL 14+). Table, source type, bytes and rows proces= sed, rows excluded and skipped, and a percentage with ETA when the total is= known: the file size for COPY FROM a server-side file, the ro= w estimate for COPY TO.
  • Health = Score from the stock exporters. The time-series mode of the Health= Score now works with plain postgres_exporter, node_expo= rter and pgbouncer_exporter, so an existing Prometheus = or VictoriaMetrics setup is enough. Metrics of Patroni clusters are mapped = automatically by scope and member name.
  • Cold ta= bles are not penalized. Tables with no writes for longer than an i= nactivity threshold (7 days by default) no longer lower the score for dead = rows and missing vacuum. A separate recommendation lists those that need a = single VACUUM (FREEZE, ANALYZE), and the Maintenance page can = filter all, active or cold tables.

Features

  • Health = Score: a composite 0-100 score per instance across eight categorie= s, with prioritized recommendations tied to a specific database. Optionally= computed from Prometheus/VictoriaMetrics time series, with a trend and a s= easonal baseline.
  • Query a= nalysis: top queries by execution time and WAL, running and blocke= d queries with wait events, and pg_stat_statements snapshots. = Snapshots can be saved to a storage database, shared by URL, and diffed aga= inst another snapshot or against live data.
  • Index a= dvisor and analysis: missing B-tree indexes proposed from pg= _stat_statements of every host, optionally backed by auto_expl= ain plans from the log. Bloat, duplicates, invalid indexes, and an a= nswer to "may I drop this index?" that accounts for statistics resets, per-= host counters and partitioned indexes rather than trusting idx_scan = =3D 0.
  • Schema = checks: 17 checks for structural defects, such as sequences close = to exhaustion, tables without a primary key or a public schema= anyone can create objects in, each with a severity and remediation SQL.
  • I/O ana= lysis built on pg_stat_io (PostgreSQL 16+), with WAL = on PostgreSQL 18.
  • Log ins= ights: log search in OpenSearch, Elasticsearch, VictoriaLogs and Y= andex Managed PostgreSQL, event categories, auto_explain plan = trees with findings, and plan regressions between two time windows.
  • Automat= ic snapshots: a separate daemon captures query and instance-wide l= ock-contention snapshots on activity spikes or primary/replica role changes= . The evidence is there the next morning when someone asks what happened at= 03:00.
  • Tables = and maintenance: size with a TOAST breakdown, per-table describe, = and autovacuum and wraparound thresholds computed with PostgreSQL's own for= mula. Also progress of VACUUM, ANALYZE, CREATE INDEX, CLUSTER, base backups= and COPY, and hot tables and indexes derived from scheduled delta snapshot= s.
  • Cluster= discovery: a static list, Patroni clusters from etcd or Consul, o= r Yandex Managed Service for PostgreSQL; databases are discovered from the = cluster itself.
  • Access = control: three auth modes (open, static API keys, or OIDC with enc= rypted session cookies), Casbin RBAC, personal access tokens and per-identi= ty rate limiting.
  • MCP con= nector: a read-only MCP server that gives AI assistants the fleet = diagnostics. It has 35 tools, including index recommendations, schema check= s, I/O, log search, plan regressions and live activity sampling. It also ha= s 5 prompts, per-user token passthrough and an embedded knowledge base of t= he scoring rules.

The interface is available in English, Russ= ian and German.

Requirements

  • PostgreSQL 14 o= r newer (tested on 14-19b) for the monitored clusters, reachable with a rol= e holding pg_monitor and CONNECT on every monitor= ed database.
  • pg_stat_s= tatements (or pgpro_stats on Postgres Pro) for query st= atistics; pgstattuple is optional.
  • Activity & = Waits needs compute_query_id set to on, or to auto with pg_stat_statements or pgpro_stats preloaded.
  • The I/O page re= quires PostgreSQL 16 or newer.
  • Optional: etcd = v3 or Consul read access for Patroni discovery.
  • Optional: a log= source (OpenSearch, Elasticsearch, VictoriaLogs or Yandex Managed PostgreS= QL) for log search and plan insights, with auto_explain enable= d for plan analysis.
  • Optional: a sep= arate PostgreSQL database for snapshots, hot-object history and personal ac= cess tokens. Everything else works without it.

Documentation

License

Dasha is licensed under the GNU General Pub= lic License v3.0; see LICENSE.

Project page: gith= ub.com/dbulashev/dasha

This email was sent to you from Dasha. It was delivered on their behalf by the PostgreSQL project. Any questions about the content of the message shou= ld be sent to Dasha.

You were sent this email as a subscriber of the pgsql-announce mai= linglist, for the content tag Related Open Source. To unsubscribe from further emails, or change which emails you want to receive, please click th= e personal unsubscribe link that you can find in the headers of this email, or visit https://lists.postgresql.org/unsubscribe/.
 
--===============7676022457510646191==--