Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
43 changes: 43 additions & 0 deletions CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -33,13 +33,35 @@ Releases before 3.0.0 are not archived: those entries carry no prose to move.
- **People can give feedback on investigation reports** ([#5926]) - A report can be marked right, partly right or wrong, with a corrected verdict, the step the agent missed and a comment, from the Viewer's Investigations window (Report Feedback tab, admin login) or the web Admin page's new Report Feedback tab (recorded only under an admin store login; list-only otherwise). `get_investigation_reports` now returns each report's feedback and, per alert family, how many investigations ended as false positives before and after people's corrections.
- **Lite: Level 1 agent investigations** ([#5820]) - when an alert fires on a notification route whose agent level is 1, Lite opens an investigation request for it, and when `mcp_allow_investigations` is on with MCP writes off it serves four tools (`list_investigation_requests`, `claim_investigation_request`, `submit_investigation_report`, `get_investigation_reports`) so a local agent can claim the request, read its brief, and file a report that Lite keeps beside the alert. Requests, reports, notes and the journal live in Lite's own DuckDB file; nothing writes to a monitored server; with both MCP settings on the investigation tools stay off and Lite logs one Critical line saying so.
- **Deadlock graph window: a visible Copy button for the SQL text** ([#5663]) - the only ways to copy a process's SQL text were the right-click "Copy SQL Text" menu item or selecting the text box and pressing Ctrl+C, and the menu was easy to miss. The Properties panel now has a Copy button on the same row as the "SQL Text" heading, in Lite and the Darling viewer. It copies the same full SQL text as the menu item, through the same code, and is only shown when the process has SQL text.
- **The web dashboard charts file I/O throughput, tempdb size and I/O, and collector duration** ([#6296], [#6297], [#6298]) - a new I/O Trends tab (up to the last 7 days) charts file I/O throughput per file, tempdb allocated size and per-file tempdb I/O latency, and Collection Health charts collector duration over time for the five slowest collectors, as the desktop viewer does.
- **The web dashboard compares Top Queries, Top Procedures and Query Store with an earlier period** ([#6300]) - each of the three grids, and the Top Queries and Top Procedures cards on the CPU tab, has a Compare picker (Yesterday, Last week, Same day last week) that shows each row's change in executions, average duration and average CPU against the earlier period, with NEW and GONE badges. Each window is its own top 100, so a row ranked below 100 in the earlier window reads NEW; the card says so.
- **The web Blocking page shows the lead-blocker tree of a blocked process report** ([#6292]) - each row has a new Block chain button that opens the lead-blocker tree of that report, as the desktop viewer does. The Darling Viewer and Lite block chain windows now show transaction start times in the chosen display mode (Server, Local or UTC); they showed the raw server clock in every mode. The web serves them in UTC.
- **The web System Events tab charts the corruption and contention counters** ([#6295]) - it now charts bad pages, dump requests, access violations, write access violations, non-yielding tasks, latch warnings and sick spinlocks over time, as the desktop viewer does.
- **The web Active Queries tab picks a stored snapshot** ([#6302]) - the tab now opens on the latest stored snapshot and has a picker over the snapshots in the window, with an "All snapshots in window" choice, and a chart's "Active queries at this time" opens on the nearest snapshot. The Darling MCP tool `get_active_queries` takes two new optional parameters: `snapshots_only` lists the window's snapshot times, newest first, with each snapshot's row count, and `collection_time` returns one snapshot's rows. Lite's `get_active_queries` is unchanged.
- **Alerts show how many days in a row they fired at the same time of day** ([#6261]) - in Darling and Lite, `get_alert_history` takes a new optional `include_streak` (default false) that adds `streak_days` (days in a row an alert fired within 30 minutes of the same time of day, counting muted fires, up to 7) to each row, and investigation briefs show the same streak beside the recurrence counts.
- **The web server page shows the Daily Health Calendar as a month calendar** ([#6294]) - the web showed the daily summary without a calendar. It is now a month calendar of daily health bands with month buttons, a day detail that gives the reasons, and buttons that open the matching tab at that day.
- **The web dashboard draws the Query Heatmap as a colour grid** ([#6293]) - the web showed the heatmap only as a table of its cells. It now draws the grid as the desktop viewer does: time across and duration buckets down, a one-hue shade by query count, a metric picker, hover details for each cell, a jump from a selected cell to Active Queries at that time, and a Show as table switch. A window up to 600 px wide opens on the table. The heatmap read returns up to 1000 cells, up from 500, and bins get wider on long windows so the grid stays near 144 columns.
- **The web Memory tab charts memory grant sizing and activity, and `get_memory_grants` can return the same series** ([#6299]) - the web Memory tab had no chart of grants. It now draws memory available, granted, used and target over the selected window, one line per resource pool. The Darling MCP tool `get_memory_grants` takes two new optional parameters, `chart` and `bucket_minutes`; with `chart` true it answers the same series, one point per bucket per resource pool. Lite's `get_memory_grants` is unchanged.
- **The web dashboard opens a per-row history for Top Queries and Top Procedures** ([#6303]) - each row has a History button that opens that query's or procedure's trend (procedures include average CPU), as double-clicking a row does in the desktop viewer. The Darling MCP tool `get_procedure_duration_trend` takes an optional `procedure_name` to read one procedure. Lite's `get_procedure_duration_trend` is unchanged.
- **The web Overview draws five strips on one time axis** ([#6291]) - CPU, waits per second, blocking and deadlocks, buffer pool and file I/O latency are drawn as five stacked strips with a shared crosshair and a drag-zoom across all five. As in the desktop viewer, a baseline band (mean plus or minus 2 deviations) sits behind CPU, waits, blocking and I/O latency, and CPU, waits and I/O points outside it get a red dot; with too little history the strip says there is no baseline yet. The Overview no longer shows its separate CPU, Memory, Blocked report and Deadlocks charts or its File I/O panel; those stay on their own tabs.

### Changed

- **Lite's Query Store regressions now show timeouts and errors beside cost, and no longer report logical reads** ([#6308]) - Lite's get_query_store_regressions averaged every run type into one figure and returned logical reads, so a query whose timeouts or errors rose could be hidden and the same tool named differed from Darling's. It now compares Regular runs per execution (weighted by execution count), shows Aborted and Exception runs side by side, lists a query for cost, timeouts or errors with a regression_reason, uses Darling's thresholds, and drops baseline_reads, recent_reads and io_regression_percent. The shared tool description changed with it.

### Fixed

- **The Darling viewer's Manage Servers window now shows the service's version, once, instead of its own version on every server row** ([#5668]) - the grid's Installed Version column filled every row with the viewer's own version, so an old viewer talking to a newer service read as up to date. The service now writes its product version and process start time to the store when it starts (a new table, `config.service_version`, one row per service host, store schema V174, created by the first start after the upgrade), and Manage Servers shows it once above the grid beside the viewer's own version: one line per service host when more than one service writes to the store. A store with no row yet, or without the table, reads `unknown (older than 3.11)`. A write that fails at start logs a warning, and the service tries it again on every sweep until it succeeds. The Installed Version column is removed. The line is display only and shows no warning.
- **A collector that recovered no longer stays at WARNING** ([#5665]) - the 7-day failure rate gave WARNING above 20% errors and a later success never cleared it, so a new target with 2 errors in its first 8 runs stayed WARNING until its tenth run. The rate now counts only once the window holds 10 runs, or fewer when a collector runs so rarely that a full 7-day window holds fewer (7 for a daily collector, so a daily collector that failed 6 days of 7 and succeeded today stays WARNING), and a collector whose newest 3 runs in a row succeeded is HEALTHY with a note "recovered: the last 3 runs succeeded; last failure <time> UTC" in the Note column of Lite and the Darling viewer, in a Recovered column of the Darling web dashboard, and as note_summary (recovered_note on the shorter rows) in the MCP get_collection_health read. The note appears only on a row the recovery rule moved to HEALTHY, never on one that another rule bands WARNING or STOPPED. A collector that alternates between failing and succeeding stays WARNING, a collector with few runs whose latest run failed stays WARNING, and the staleness, abandon-rate and lost-databases rules are unchanged. The fleet overview card and the Darling viewer's Overview cards agree: when a collector's failure rate alone would band it WARNING, the roll-up reads that collector's newest 3 runs in one follow-up statement (none when no collector needs it), so a recovered collector reads HEALTHY on the fleet card as on its own tab and an alternating one stays WARNING.
- **The upgrade and installer accept a hand-named `darling.json.bak-*` backup in the install folder** ([#5659]) - In 3.10.0 the check that runs before the lock refused an install folder holding a backup such as `darling.json.bak-before-edit`, because the service owns every `darling.json.bak-*` file after a start and the check trusted it only on the stamped names the CLI writes. The refusal said the folder was writable by ordinary users and told the operator to move the whole install. The check now trusts the service account on every file in the install root whose name starts with `darling.json.bak-`, the same prefix the hardening uses, and nowhere else. Any other grantee on such a file, the service on another file, a folder with that name, a link, a file with more than one name (a hard link) and a file below the root still refuse. A backup-named hard link is now reported as a finding of its own, and the lock and the install pass leave it alone, so it never gets the service as owner. The lock also leaves any other file with more than one name alone, wherever it is in the install tree, and lists it as still open instead of changing its owner or resetting its permissions, because either would change the file's other name too. That owner step opens the install folder once and changes an owner only on an object whose real path it confirms is inside that folder, lists a name that ends in a dot or a space instead of acting on it, and turns on SeTakeOwnershipPrivilege and SeBackupPrivilege for itself and puts them back as it found them. When a refusal lists only backup files, it now starts with a note that names them, says they can hold the secrets that were in darling.json and can be moved to a folder only Administrators can read, and then gives the usual advice unchanged. The check, the install-time hardening, the upgrade's manifest guard, the service sweep and the CLI check now match one prefix per language by ordinal, case-insensitive name, so a Turkish-locale host reads the names the way the service does. The CLI permission check now lists the backups of the config it was given (`prod.json.bak-*` for `prod.json`) instead of always `darling.json.bak-*`. A name that ends in a dot or a space is not treated as a backup by the check, the install-time hardening, the service sweep or the CLI check, because Windows resolves such a name to a different file.
- **The installer and the upgrade now compare a folder the same way whether it is typed as a short path, with a trailing backslash or in another case** ([#5682]) - the same-folder and at-or-under checks (the folder the upgrade runs from, the source folder, the service's registered folder, the user-profile check on the install folder, the `file:` reference check on a `dataDirectory`, `pfxPath`, `certPath` or `keyPath` value, and the refusal advice) now put both paths in one long spelling first. A folder given as an 8.3 short path, with a trailing backslash or in another case gives the same answer as its long form.
- **Default trace events no longer time out on small servers** ([#6311]) - The default trace collector sorted its result on the server, and that sort asked for a 57-75 MB memory grant to return a few dozen rows. A server with less query workspace memory than that left the collector waiting for the grant until its 60 second command timeout, so the cycle failed with "Execution Timeout Expired" and collected nothing. The query no longer sorts (the collector orders the rows itself, newest first, as before) and runs single-threaded, so it requests about 1.5 MB.
- **Azure SQL Database: a connection outage is no longer logged as a missing XE session, and master is no longer a failed database** ([#6312]) - with every user database refusing connections during a firewall outage, the `attention` row was logged as a missing XE session with master's "session is not running" text; the connection failure now wins. The `attention` and `blocked_process_report` collectors skip master, so it no longer counts as a failed database. The blocking snapshot tests for a user process with `is_user_process`, so system sessions above 50 stay out of it.
- **MCP answers no longer misstate sub-millisecond CPU, the monitor's own xmin, the tag window minimum or a two-session move** ([#6305]) - `analyze_server` advice shows per-execution figures under 1 ms in microseconds instead of "0 ms"; `get_pg_xmin_horizon` no longer reports a holder at xmin age 0 (the monitor's own session); `get_resource_use_by_tag` says its query_store section needs hours_back 24 or more; `compare_analysis` bands a near-flat baseline against a noise floor (5% of the mean, 2 sessions for session counts), so a two-session move is no longer "worse" at 25 sigma.
- **The Admin Agent changes list names the server, Admin tabs wrap, and the server page header stays current** ([#6310]) - the Agent changes list showed "Server <id>" instead of the server's name, and its Kind column showed the raw `session_kill` where it now reads "kill session". The Admin tab bar wraps so every tab is readable at 1280 px, and the server page header tiles no longer lag the charts after a tab click. The Server actions switches lock with a "Saving" line while a change is in flight; before, a click during the change was silently dropped.
- **The web dashboard's sidebar no longer takes 260 px of a phone-width window** ([#6315]) - at 768 px and narrower the sidebar becomes an overlay menu opened from a new top bar, and the page content takes the full width with a 16 px side gutter. The fleet and Availability Group tiles wrap when their column is narrow, long selects shrink to their card, and the footer wraps.
- **The plan regression finding now also catches queries whose runs time out or fail** ([#6309]) - The finding counted only runs that completed, so a query that started timing out or erroring while its successful runs stayed cheap was never reported. It now flags a query whose timeouts or errors in the last 24 hours rise to at least twice the earlier rate (or appear where there were none, with at least 3 runs), in Darling and Lite, says so in its advice, and lists those queries with their recent and baseline timeout and error counts in the finding's drill-down next to the cost offenders.
- **The Query Store collector no longer joins Query Store views in one statement** ([#6257]) - the payload read and the text fetch now read each Query Store view once into its own temp table and join the temps, so a primary no longer merges each view's in-memory data in the middle of a join. The collected rows are unchanged. REF: [#6257]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6257

## [3.10.0] - 2026-10-09

Expand Down Expand Up @@ -5237,3 +5259,24 @@ Full entries: [docs/changelog/3.0.md](docs/changelog/3.0.md)
[#5922]: https://github.com/erikdarlingdata/PerformanceMonitor/pull/5922
[#5925]: https://github.com/erikdarlingdata/PerformanceMonitor/pull/5925
[#5926]: https://github.com/erikdarlingdata/PerformanceMonitor/pull/5926
[#6308]: https://github.com/erikdarlingdata/PerformanceMonitor/pull/6308
[#6311]: https://github.com/erikdarlingdata/PerformanceMonitor/pull/6311
[#6312]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6312
[#6305]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6305
[#6310]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6310
[#6296]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6296
[#6297]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6297
[#6298]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6298
[#6315]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6315
[#6309]: https://github.com/erikdarlingdata/PerformanceMonitor/pull/6309
[#6300]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6300
[#6292]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6292
[#6295]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6295
[#6257]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6257
[#6302]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6302
[#6261]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6261
[#6294]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6294
[#6293]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6293
[#6299]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6299
[#6303]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6303
[#6291]: https://github.com/erikdarlingdata/PerformanceMonitor/issues/6291
Loading