ADQM Control releases

Introduced two levels of query grouping: families of queries and query variants within a family

Added the ability to leave comments on query groups

Introduced the query queue model including query service time and time spent waiting in the queue

Introduced the User load graph

Introduced the Duration delta graph with details for the selected time window

Added cards with statistics on table queries on the page with ADQM table details

Introduced table revisions and column revisions on graphs

Added the Groups option to the Top 10 queries dashboard section

Implemented treemap charts to visualize data in the Top 10 tables (Aggregated) and Top 10 queries (Groups) sections on the Dashboard page

Introduced brute force protection on login

Added the ability to block/unblock users

Added the ability to delete users

Introduced the Online user status with displaying active sessions

Introduced an alert about an unsupported ADQMDB version

Added support for Ubuntu 24.04 LTS

Added support for Astra Linux SE 1.8.6 Orel and Voronezh

Reworked calendar and time range picker of the Time control

Reworked the Refresh control: period now follows the selected time window

Significantly sped up building graphs for queries over long time windows

Significantly sped up working with tables and table groups over long time windows

Changed the default time range to the period from the start of the current day until now

Reworked range presets

Added sign out of other sessions on password change

Added backoff for hosts with permanent ClickHouse errors

Added estimation of persistent metrics for big tables

Fixed empty lists of queries and hosts while metrics exist in the past

Fixed the Requests imbalance graph ignoring idle tables

Fixed table group summary and graphs ignoring the host filter set in the table list

Fixed cleanup resetting aggregation bounds and triggering a full rollup rebuild

Fixed the calculation of table lifetime

Fixed lost initial request on a background tab

Fixed data race in collection of hosts

Upgraded Go to version 1.26

Upgraded TimescaleDB to version 2.29 with TimescaleDB Toolkit 1.19

Added text-free metric rollups

Reworked aggregation into a cascade pipeline

Added rollup integrity job and aggregation bounds integrity check

Added table aggregation job with per-host bounds

Replaced advisory locks with a resource locker

Added batched cleanup deletes with per-table tuning

Made the schedule of host checks configurable

Added switch to disable recomputation of persistent metrics

Added log rotation by interval

Added check window to integrity checks and Maintenance CLI

Added column codec cleanup job with retention setting

Implemented query duration comparison backend

Added logging of panics in HTTP servers

Added closing database connections after migration

Introduced the Host graph on the Queries page — distribution of queries across cluster hosts with anomaly detection

Introduced table groups (aggregated tables)

Added alerts about host and ADQMDB service unavailability

Added data cleanup/retention settings

Added the aggregated view to the Top 10 tables dashboard section

Added a modal window with graphs for top 10 tables on the Dashboard page

Added information about the lifetime of tables

Added graphs related to request count for tables

Added support for RHEL 9

Added support for RED OS 8.0 Certified edition

Added the ability to select the level of detail (Low/High) for graphs by tables and queries

Moved alert cleanup settings to the Settings → Cleanup page

Added support for non-UUID query IDs

Corrected the alert cooldown period check

Long-running migration now receives a stop signal on cancellation instead of leaking resources

Fixed an error related to stopping the collection of ADQMDB metrics when Prometheus was unavailable while the host was still in the cluster

Query monitoring now tracks rows and bytes processed per query and per aggregated bucket

Added basic authentication support for adqmc-alert-receiver

Bundle contract version is now 2.1

ADQM cluster topologies can now be determined via ZooKeeper as an alternative to Prometheus

Implemented the pull/push option for the job-log command of ADQM Control Maintenance CLI

Removed C++ services

Performed internal refactoring

Minimum version of ADQMDB is now 25.3.2.39

Updated permissions in the role-based access model of ADQM Control

Added support for filtering queries by multiple exception codes

Added support for filtering queries by exception code presets

Implemented various backend and frontend optimizations

Introduced ADQM Control Maintenance CLI

Removed the Zookeeper service

Added support for Alt Linux 10.2

Added support for Astra Linux 1.7.6 Voronezh

Updated submodules

Increased the minimum version of ADCM to 2.7.1

Introduced the Duration, Kind, and User graphs for queries

Introduced incidents

Added support for SSL in all services of ADQM Control

Added the Comment field for ADQM table columns on the Table details page

Implemented a new alert on missing the system.query_log table

Added the threshold label to ADQM Control alerts

Implemented acknowledgement for alerts

Added multi-user, query duration, query text, and exception filters on the Queries page

Added search filters on the Users page

Extended Heat map with statistics information for each host

Refactored paginator on the Queries page by removing the page counter

Introduced ADQM Control services, including Agents, Alert generator, Alert manager, Alert receiver, and Backend

Introduced basic functionality for managing users within ADQM Control

Introduced the Dashboard page providing an overview of the physical cluster state, alerts, top 10 tables, and top 10 queries

Introduced the Tables page for displaying table information over a specified time window

Introduced functionality for viewing detailed information about columns within tables over a specified time window

Introduced charts for tracking table metrics over a defined time period

Introduced charts for tracking column-level metrics over a defined time period

Introduced the Queries page that displays queries providing sorting and filtering options

Introduced the Alerts history page that displays a full list of alerts providing sorting and filtering options

Introduced the Heat map view for alerts — a visual representation of alert distribution over time

Introduced the Settings page for configuring ADQM Control parameters, such as password complexity and rules for generating alerts

Introduced functionality for using the ADPG service or an external PostgreSQL instance as a database for ADQM Control

Introduced functionality for using either the built-in Zookeeper service or an external ZooKeeper instance as a coordination service