Fixed:
- Cancel SQLServer Agent history job on check cancellation (#22753)
Added:
- Add
enable_legacy_tags_normalizationoption to preserve hyphens in tag values when set to false. (#22303) - Require key_prefix for SQLServer connections to avoid unsafely sharing connections across threads (#22557)
Fixed:
- Clarifies the default ADO provider in conf.yaml.example (#21364)
- Updates collect_raw_query_statements config description (#22611)
- Fix AgentJobs, Deadlocks and XE collections to run async by default (#22633)
Fixed:
- Fix AgentJobs, Deadlocks and XE collections to run async by default (#22633)
- Cancel SQLServer agent history job on check cancel (#22753)
Fixed:
- Fixed an issue where schema collection thread connections could get abruptly closed from the main check loop (#22471)
Fixed:
- Improve SQL Server USE stmt identifier escaping (#22277)
- Fix STRING_AGG truncation error for tables with many columns by casting to NVARCHAR(MAX) (#22338)
- Fix KeyError in statement metrics when available metric columns change between collection cycles. (#22318) Optimize StatementMetrics cache to only store fields we need to compute statement metrics. (#22358) (#22376)
Fixed:
- Replace SQLServer checks for year with checks for major version. (#22285)
- Fix default check in SQL Server major version (#22291)
Fixed:
-
Fix table row count calculation for sqlserver.table.row_count. Previously, we joined with indexes which produced one row per table-index combination before grouping. This resulted in inflated row count values when grouping by table because the multiple index rows were summed together.
The fix is to conditionally sum the row count: only when the index_id is in (0, 1), which translates to heap and clustered index partitions. Both of these contain the true row count for the given partition, whereas other partition types (e.g. non-clustered indexes) should not be considered in the row count calculation. This leads to the expected value being calcualted. (#22055)
Added:
- Migrate SQL Server to new schema collector, which provides improved performance in the Agent and allows the backend to handle larger schema collections (#21734)
- Add DBM Agent health events to SQLServer, including basic initialization checks, unhandled errors, and missed collections (#21868)
- Upgrade base version for Postgres, MySQL, and SQLServer (#21906)
Fixed:
- Support
only_custom_queriesconfiguration option in the SQL Server integration (#21304)
Added:
- Add additional context to the database_identifier configuration description. (#21575)
Fixed:
- Remove ddagenthostname from metrics for Postgres, MySQL, and SQLServer (#21523)
Added:
- Bump Python to 3.13 (#21161)
- Add SQL Server database metric
user_access. (#21406) - Normalize
sqlserver_servernameandsqlserver_instancenametag values to avoid duplicate tags. (#21475) - Bump datadog-checks-base to 37.21.0 (#21477)
- Add
replace_bind_parameterto sqlserver obfuscator option to support obfuscate bind parameters like@P1. (#21481)
Fixed:
- Compile and reuse connection error regex patterns (#21319)
Fixed:
- Remove ddagenthostname from metrics for Postgres, MySQL, and SQLServer (#21523)
Added:
- Update core tags with agent host name for DB integrations(postgres, mysql, sqlserver) (#20991)
- Enable collect_settings by default for DBM enabled integrations (#21172)
- Update dependencies (#21217)
Added:
- Add new collect_* configuration options (collect_schemas, collect_deadlocks, collect_xe) to replace deprecated *_collection options while maintaining backward compatibility. (#20599)
Fixed:
- Add deprecated
collect_schemas,collect_xeandcollect_deadlocksoption toconfig.yaml.example. These config options are deprecated at Agent 7.70.0 and will be removed in a future release. (#20781) - Lift
pyodbcdependency exclusion for macOS on AArch64/ARM64 (#20812) - Improve descriptions and examples in example configuration file (#20878)
Fixed:
- Fixes "Column 'sys.foreign_keys.object_id' is invalid in the select list ..." error with the
FOREIGN_KEY_QUERY_PRE_2017query (#20546) - Emit
session_idin XE events as an integer always, remove unnecessaryevent_sourcefield (#20559) - Only submit plan events when we have a plan definition (#20584)
- Remove relative imports for non parent modules (#20646)
Added:
- Add $full_server_name support as template variable to sqlserver template config key (#20342)
- Add size and row count collection for tables in SQLServer (#20367)
- Update dependencies (#20399)
- Add
sys.dm_exec_sessions.client_interface_nameto activity samples payload. (#20463)
Fixed:
- Update SQL server to TagManager and fix race conditions in tag values dependent on server queries (#20424)
Added:
- Fill in missing fields for XE events (#20293)
Fixed:
- Add object_name to rpc_completed XE events, add activity ID and activity ID Xfer to error events (#20302)
Added:
- Add database and Azure name to SQL Server database identifier template variables (#20301)
Added:
- Enable HA agent support for DBM integrations (#20124)
- Added SQLServer Extended Event Handlers (#20229)
Fixed:
- Use DB_NAME for query metrics in an Azure SQL database since joining on sys.databases doesn't always work. The dbid from dm_exec_query_stats doesn't match the one in sys.databases. (#20097)
Added:
- Create exclude_hostname option for Postgres, MySQL, and SQLServer (#20094)
Added:
- Added a new configuration option
database_identifier.template. Use this template to specify the unique identifier for a database instance, separate from the underlying host. Theempty_default_hostnameconfiguration option is now respected and will omit thehosttag from database instances when enabled. (#19341)
Added:
- Optimize SQL Server integration by removing redundant query comment parsing using
extract_sql_commentsand switching stored procedure name extraction to useOBJECT_NAME(sproc_object_id, dbid); whendisable_secondary_tagsis enabled, fall back to parsing the procedure name from the raw query text. (#19935) - Update dependencies (#19962)
Fixed:
- Remove unnecessary
likefrom SQL Server deadlock query (#19921)
Fixed:
- Fix SQL Server integration when database is using case sensitive collation (#19930)
- Fixed support for FreeTDS driver (#19931)
Changed:
- Optimize sqlserver index usage stats query (#19807)
Added:
Fixed:
- Bump base check dependency to use new assert method functionality. (#19763)
- Fix a bug where
tempdbis wrongly excluded from database files metrics due to all instances inherited fromSqlserverDatabaseMetricsBaseshare the same reference of auto-discovered databases. (#19803)
Added:
- Add support for collecting raw query statements and explain plans when
collect_raw_query_statement.enabledis true. (#19421) - Add
raw_signatureto raw query plan event payload. (#19495) - Add deadlock support for Azure DB (#19577)
- Add SQLServer ServerName and InstanceName to tags (#19600)
Added:
- Update dependencies (#19430)
Fixed:
- Fix KeyError in SQL Server schema collection caused by case-insensitive database name mismatches. (#19384)
- Bump datadog-checks-base version (#19478)
Changed:
- Fall back to
system_health/event_filewhen querying deadlocks ifdatadogXE session wasn't created. (#19189)
Added:
- Update configuration structure and allow configuration of all database metrics (#19111)
- Send schema name as part of index usage metrics (#19266)
- Add schema tag to db_fragmentation metrics for sqlserver (#19277)
Added:
-
Submit database_hostname with database instance and metrics for MySQL, Postgres, and SQLServer (#18969)
-
Add lookback_window config parameter to query_metrics.
The current lookback window defaults to 2 times the collection interval, and is not able to be overridden. This means that infrequently-run queries are unlikely to have metrics captured for them. One common use case that falls into this bucket is ETL queries which can run hourly or even daily. These have a very small chance of having metrics captured for them. In that case, we will support setting a lookback window that will include such queries. (#18979)
-
Add config option
azure.aggregate_sql_databasesto report multiple azure sql databases as one database host. This is an opted in feature and is disabled by default. (#19032)
Fixed:
- Fix missing appended SQL comments. (#18958)
- Fix
azure_sql_server_databaseresource tag to use Azure SQL Database{fully_qualified_doman_name}/{database_name}. (#19014) - Update SQLServer agent jobs metrics to be DBM only. (#19033)
- Fix duplicate deadlock events (#19139)
- Fix poor query signature correlation for deadlocks. (#19142)
Fixed:
- Use alternative schema collection query for sqlserver 2016 and older due to STRING_AGG not being supported until SQLServer 2017 (#19110)
Added:
- Add
serviceconfigured in integration init_config or instance config to the DBM events payload. The configuredservicewill be converted to tagservice:<SERVICE>and applied to query metrics, query samples and explain plans. (#18846) - Add extended events (XE) session monitoring (#18875)
- Migrate following dynamic metrics to database_metrics for better maintainability and testability.
- SQLServer AlwaysOn metrics
- SQLServer FCI metrics
- SQLServer file stats metrics
- SQLServer primary log shipping metrics
- SQLServer secondary log shipping metrics
- SQLServer server state metrics
- SQLServer tempdb file space usage metrics
- SQLServer index usage metrics
- SQLServer database index fragmentation metrics
- SQLServer os tasks metrics
- SQLServer master files metrics
- SQLServer database files metrics
- SQLServer database stats metrics
- SQLServer database backup metrics
- SQLServer os schedulers metrics
- SQLServer database replication stats metrics
- SQLServer availability replicas metrics
- SQLServer availability groups metrics Increase database backup metrics and index fragmentation metrics collection interval to 5 minutes. (#18883)
Fixed:
- Prevent hostname evaluating to None in sqlserver check (#18237)
- Update sqlserver metric collection row_key to prevent overwriting metric information for two queries with the same query_plan_hash but part of two different stored procedures. (#18882)
Changed:
- Use
datadogXE session as the default for deadlock monitoring. Fall back tosystem_healthif unavailable. (#18781)
Removed:
- Remove support for Python 2. (#18580)
Added:
- Added deadlock collection feature to the SQL Server integration. (#18108)
- Bump the python version from 3.11 to 3.12 (#18207)
- Add the
propagate_agent_tagssetting. When set totrue, the tags from the agent host are added to the check's tag for all instances. (#18557)
Fixed:
- Bump the version of datadog-checks-base to 37.0.0 (#18617)
Fixed:
- Updated SQL Server Agent job query for completed jobs, significantly reducing query times for large job history logs (#18760)
Changed:
- Bump minimum version of base check (#18733)
Added:
Fixed:
- Fix ODBC config handling for Linux (#18586)
Fixed:
- Fix ODBC config handling for Linux (#18586)
Fixed:
- Bump lxml to 4.9.4 (#18244)
Fixed:
- Directly collect the time offset for converting times in SQL Server Agent tables to epoch timestamp through a single, deterministic function. This method avoids calculating the offset as a difference between the server time and UTC time, which led to inaccurate offsets in some cases. (#18292)
Added:
- Added collection of SQL Server Agent jobs history and activity (#17787)
Fixed:
- Fix password obfuscation in OLE DB driver error message when one or more backslashes exist in the password. (#18203)
Fixed:
- Revert "[dbm] fix flapping sqlserver_version" (#18249)
Added:
- Adding schema collection to sqlserver Schema data includes information about the tables, their columns, indexes, foreign keys, and partitions. (#17258)
- Set sqlserver integration default odbc driver to agent embedded msodbcsql18 on Linux platform (#17729)
- Update embedded odbc driver config odbcinist.ini to use msodbcsql 18.3.3.1 (#17765)
- Update dependencies (#17817)
Fixed:
- Filter query metrics by database_name if check is configured to a single non-master db (#17717)
- [sqlserver] fix missing sqlserver_version (#17750)
- Hyphen in database name support (#17775)
- Upgrade
azure-identitydependency (#17862) - Move sqlserver schema foreign key information to the dependent table. (#17933)
Added:
- Update dependencies (#17424)
Fixed:
- Fix Always-On metrics query for replica_failover_mode and replica_failover_readiness. Previously the query returned a cartesian product of rows which could result in incorrect metrics in some cases. (#17503)
- Emit database_instance metadata before collecting metrics Decreased database instance collection interval from 1800 seconds to 300 seconds to improve reliability (#17675)
Added:
- Add
sqlserver.database.replica.transaction_delaymetric to report database level replica transaction delays (#17345) - Add a new config option include_db_fragmentation_metrics_tempdb to skip index fragmentation metrics for tempdb. By default, index fragmentation metrics are NOT be collected for tempdb. (#17361)
- Migrate
SQLServer index usage metrics,SQLServer database index fragmentation metricsandSQLServer database backup metricstodatabase_metrics. IncreaseSQLServer database index fragmentation metricsandSQLServer database backup metricsdefault collection interval to 5 mins. (#17374)
Fixed:
- Break circular dependency between sqlserver check and connection (#17275)
- Fix missing file stats metrics
sqlserver.files.write_io_stall_queuedandsqlserver.files.read_io_stall_queuedfor sqlserver newer than 2012 and not Azure SQL Database/Azure SQL Managed Instance (#17299) - Remove db exist check if ignore_missing_database is set to false. In this case throw SQLConnectionError instead of ConfigurationError when database does not exist. (#17300)
- Fix zero value bug for
sqlserver.replica.transaction_delaymetric due to integration wrongly pulling transaction delay performance counter fromSQLServer:Database Mirroringinstead of the the correct object_nameSQLServer:Database Replica(#17345) - Fixed a bug where specifying
nullforcustom_metricswould cause the SQL Server integration to crash (#17430) - Prevent agent from unsupported USE commands on Azure SQL DBs (#17448)
Added:
- Update custom_queries configuration to support optional collection_interval (#16957)
- Tag SQL Server agent queries with service:datadog-agent (#17162)
Fixed:
- Skip index usage collection for tempdb, fixes a blocking query (#16977)
- Fix an issue where logging of obfuscation errors would still hide the statement when it's part of a stored procedure. If the stored procedure obfuscation fails, the
procedure_signaturetag will be set to__procedure_obfuscation_error__. (#17020) - Skip
sqlserver.latches.latch_wait_time metricon SQL Server version 2012 and 2014. (#17063) - Update the configuration to include the
metric_prefixoption (#17065) - Improve service check of autodiscovered databases for the SQL Server integration. The fix prevents agent from creating a connection per database. The fix is not applied for Azure hosted databases due to the restriction in switching to dabases within the connection. (#17139)
- Update performance counter base name query to lookup by lowercase counter name (#17207)
- Refactor SQLServer instance connector and adoprovider initialization from configuration.
The connector/adoprovider config takes precedence in the following order:
- instance config
- init config
- the default connection/adoprovider (#17245)
Fixed:
- Deal with absence of sys.database_files SpaceUsed attribute on Azure SQL Servers (#16910)
Removed:
-
Disables SQL Server metrics on Azure SQL Database instances that rely on cross-database queries. The disabled metrics are:
- sqlserver.tempdb.file_space_usage.free_space
- sqlserver.tempdb.file_space_usage.version_store_space
- sqlserver.tempdb.file_space_usage.internal_object_space
- sqlserver.tempdb.file_space_usage.user_object_space
- sqlserver.tempdb.file_space_usage.mixed_extent_space
- sqlserver.database.backup_count (#16658)
Added:
- DBM integrations now defaulted to use new go-sqllexer pkg to obfuscate sql statements (#16681)
- Bump dependencies (#16858)
Fixed:
- Improve performance of index_usage_stats query, set default collection interval to 5 minutes and allow interval to be customized. (#16645)
- Require base check that has fix for null characters in query strings (#16750)
Fixed:
- Deal with absence of sys.database_files SpaceUsed attribute on Azure SQL Servers (#16910)
Fixed:
- Replace embedded null characters with empty string in query text (#16742)
Removed:
- remove pyro4 and serpent dependencies (#16269)
Changed:
- Always use the database instance's resolved hostname for metrics regardless of whether dbm is enabled or not. For non-dbm customers, this change will result in the host tag having a different value than before. It is possible that dashboards and monitors using the integration's metrics will need to be updated if they relied on the faulty host tagging. (#16207)
Added:
- Bump the Python version from py3.9 to py3.11 (#15997)
- Add stolen memory performance counter (#16220)
- Collect sqlserver.database.files.space_used, and sqlserver.database.is_read_only/is_in_standby metrics. (#16419)
- add new obfuscator options to customize SQL obfuscation and normaliza… (#16429)
- Add ODBC driver support for the SQL Server agent (#16431)
- Add config option to set character limit of stored procedure text (#16462)
- Adds whether a SPID is a user or system process to DBM activity events. (#16465)
Fixed:
- Ensure accurate collection of query activity with statement when stored procedure is failed to obfuscate (#16455)
- Skip reporting SqlFractionMetric when base_name is None (#16469)
Fixed:
- Revert "report sql obfuscation error count (#15990)" (#16439)
- fix unexpected exception when reporting sqlserver statements obfuscate xml plan error (#16461)
Added:
- Add support for log shipping monitoring on primary and secondary instances through the
include_primary_log_shipping_metricsandinclude_secondary_log_shipping_metricsconfiguration options. (#16101) - Add obfuscation_mode config option to allow enabling obfuscation with go-sqllexer (#16125) (#16125)
- Move check config to SQLServerConfig class (#16130)
- Add unit tests to assert every configurable metrics collection (#16136)
- Updated dependencies. (#16154)
Fixed:
- Fix
aarch64compatibility of thesqlservercheck by downgradinglxmlto version 4.9.2 (16080) (#16080) - Stabilizes the
hosttag on thesqlserver.can_connectmetric and adds theconnection_hosttag on the same metric. (#16114) - [SQL Server] - Fix query stats when a query is found in more than one procedure (#16141)
Added:
- Add support for reporting SQL obfuscation errors (#15990)
- Emit SQLServer metrics queries operation time (#16066)
Fixed:
- Properly decode query_hash when statement_text is None (#15974)
- Strip sql comments before parsing procedure name (#16004)
- Bump the
pyodbcversion to 5.0.1 (#16041) - Fix config option
dbm_enabledtype to ensure it is a boolean (#16078)
Fixed
- Fix
aarch64compatibility of thesqlservercheck by downgradinglxmlto version 4.9.2 (16080)
Fixed:
- Properly decode query_hash when statement_text is None (#15974)
Fixed
- Set stored procedure collection to
run_sync=False, fixing the delay in collection of other SQLServer telemetry. (#15967)
Changed:
- Cache performance counter type to prevent querying for the same counter multiple times (especially for the per-db counters) (#15714)
- Add the
dbtag to every metric also usingdatabaseordatabase_namefor consistency (#15792) - Updates the namespace for version store performance metrics added in (#15879) to be
sqlserver.transactions.xyz. (#15904) - Updates metric documentation in
metadata.csvto add performance counter information when applicable, and provide a more complete description of available tags per metric. (15840)
Added:
- Only collect
SqlDbFileSpaceUsagemetrics fortempdb(#15906) - Add TempDB version store performance counters (#15879)
- Add TempDB page counts metrics (#15873)
- Add
index_nametag to.database.avg_fragmentation_in_percent,.database.fragment_count,.database.avg_fragment_size_in_pagesmetrics. Also add a new metricsqlserver.database.index_page_count, tagged bydatabase_name,object_name,index_idandindex_name. (#15721) - When DBM is enabled, starts collecting stored procedure metrics from sys.dm_exec_procedure_stats at 60s interval (configurable). Also adds the corresponding
procedure_metricssection to the config file. The new DBM-only metrics aresqlserver.procedures.count,sqlserver.procedures.time,sqlserver.procedures.worker_time,sqlserver.procedures.physical_reads,sqlserver.procedures.logical_reads,sqlserver.procedures.logical_writesandsqlserver.procedures.spills. (#15805) - Add additional SQL Server performance counter metrics (#15818)
- Add Index Usage Metrics for SQL Server (#15905)
Fixed:
- Restore the logic for the lookback time in the statement metrics query. It was previously the connection interval * 2, but was removed during a refactor. (#15857)
- Fix type
bytesis not JSON serializable for dbm events (#15763) - Fix sqlserver file stats metrics for Azure SQL DB (#15695)
- Fix calculation for performance counters that require a corresponding base counter type which were previously emitting a constant 100% value (such as
sqlserver.buffer.cache_hit_ratio). (#15846)
Changed:
- Collect both DBM active sessions and blocking sessions which are sleeping. See (#14054)
- Remove python 2 references from SQL Server integration (#15606)
Added:
- Add support for sending
database_instancemetadata (#15562) - Update dependencies for Agent 7.48 (#15585)
- Support Auth through Azure AD MI / Service Principal (#15591)
Changed:
- Bump the minimum base check version (#15427)
- Truncate procedure_text in sqlserver activity query (#15295)
Added:
- Add support to ingest sys.configurations for SQL Server instances (#15496)
- Update generated config models (#15212)
Fixed:
- Allow for collection of AO metrics for azure sql db (#15508)
- Fix types for generated config models (#15334)
Changed:
- Add a limit on the number of queries in the query metrics payload at the Agent (#15139)
- Do not fetch the full procedure_text in the SQLServer Qstats query (#15105)
Added:
Fixed:
- sqlserver: remove unused
procedure_textfrom metrics payloads (#15097) - SQL Server query metrics: avoid sending procedure text twice (#15091)
- Move cancel waiting logic to test functions for DBMAsyncJob (#14773)
- Bump Python version from py3.8 to py3.9 (#14701)
Fixed:
- Support SQL Server file metrics for Azure SQL DB / MI (#14679)
Fixed:
Added:
- Add sqlserver engine editition && version to DBM event payloads (#14499)
- [Unified Instance Tagging]: Add cloud_metadata to DBM event payloads (#14460)
- Send resource_type/name for sqlserver integration metrics (#14320)
Fixed:
- Exclude model from autodiscovery by default (#14493)
- Rename azure.name configuration key to azure.fully_qualified_domain_name (#14536)
- Properly initialize resolved_hostname so it is set consistently across dbm and default integration metrics (#14479)
- Restore DBM-APM link functionality for SQL Server activity sessions (#14453)
Added:
- Add resolved_hostname to metatdata (#13639)
Fixed:
- Fix a typo in the
disable_generic_tagsoption description (#14246) - Increase command_timeout default from 5s to 10s to prevent timeouts on instances with large SQL caches (#14195)
- Do not double emit instance metrics when autodiscovery is enabled (#14115)
Fixed:
- Increase command_timeout default from 5s to 10s to prevent timeouts on instances with large SQL caches (#14195)
Fixed:
- Do not double emit instance metrics when autodiscovery is enabled (#14115)
Changed:
- Lower the frequency of query metrics collection (#14033)
Added:
- Collect program_name for SQL Server activity (#13953)
Fixed:
- Fix sqlserver.database.state not being sent for all databases (#13735)
Fixed:
- Bug Fix: Azure SQL DB database name tags properly applied on perf metrics (#13757)
Fixed:
- Update dependencies (#13726)
Added:
- Add well formatted agent errors on common connection issues for SQL Server (#13436)
- Add procedure name on query metrics/events (#13484)
Fixed:
- Fix Query Metrics collection bug where long running queries were missed (#13335)
Fixed:
- Revert "Fix exception thrown when database is null in config" as it was found to introduce a regression (#13446)
Removed:
- Remove sqlserver tag truncation for metrics (#13211)
Added:
Fixed:
- Fix exception thrown when database is null in config (#12882)
- Fix Query Metrics query to correct for over-inflated / incorrect SQL Server metrics (#13089)
- Always cast provided port to str (#13055)
Fixed:
- Fix Query Metrics query to correct for over-inflated / incorrect SQL Server metrics (#13123)
Fixed:
- Allow users to configure the port as a int or as a string (#13061)
Changed:
- Use statement_start_offset to extract SQL text being run from Procedure text (#12613)
Security:
- Bump
lxmlpackage (#12663)
Added:
- Add AlwaysOn Availability Group replica status metric (#12494)
Fixed:
- Dependency updates (#12653)
- Check for port provided in config (#12610)
- Improve failed connection error messages (#12533)
- Fix documentation for GA DBM support (#12512)
- Support version specific AlwaysOn metrics (#12424)
Fixed:
- Fix engine edition logic bug and simplify configuration for Azure SQL Database (#12397)
- Support virtual file stats on SQL Server 2012 (#12094)
- Improve handling of encrypted stored procedures (#12060)
Changed:
- Remove execution plan
user_nameattribute (#12007)
Added:
- Add AlwaysOn metrics (#11979)
- sqlserver: Add option to keep aliases in sql server (
keep_sql_alias) (#12020) - Add support to ingest cloud_metadata for DBM host linking (#11982)
- Add static server OS metrics (#11864)
Fixed:
- Don't use connection resiliency for older versions of sqlserver and update valid driver list (#12026)
- Fix uncommented parent options (#12013)
- Upgrade dependencies (#11958)
- Fix missing object_name for index fragmentation metrics (#11986)
Fixed:
- Fix activity and plan host reporting (#11853)
Fixed:
- Update base version (#11826)
Added:
- Upgrade dependencies (#11726)
- Add metric_patterns options to filter all metric submission by a list of regexes (#11695)
- Add request_status in sqlserver activity query (#11699)
- Add SQL metadata to SQL Server activity events (#11689)
- Add a log line when activity limit is reached (#11661)
Fixed:
- Gracefully handle inaccessible database for
sqlserver.files.*(#11711) - Fix SQL Server false configuration error (#11664)
Added:
- Enable SQL metadata collection by default (#11606)
Fixed:
- Include SQL metadata in FQT (#11641)
Added:
- Add execution_count and total_elapsed_time fields to SQLServer Samples (#11652)
Added:
- Add missing wait_resource column for activity collections (#11638)
Fixed:
- Fix service check failures on auto discovered dbs failing full check execution (#11563)
- Use remote hostname on all metrics when DBM is enabled (#11634)
- Fix SQLServer activity query, only query for load not transactions (#11629)
Added:
- Add
pyproject.tomlfile (#11437)
Fixed:
- Fix namespace packaging on Python 2 (#11532)
- remove aggregation by SQL text from query stats query (#11524)
Added:
- Add ddagenthostname to dbm-metrics payloads (#11232)
- Add recovery_model_desc tag for sqlserver database metrics (#11210) Thanks lowlydba.
Fixed:
- Disable modified rowcounts in result sets for all connections (#11486)
Fixed:
- Update base version (#11287)
Fixed:
- Fix license header dates in autogenerated files (#11187)
Fixed:
- Bump base package dependency (#11115)
Changed:
- Add
serverdefault group for all monitor special cases (#10976) - use read uncommitted isolation level to remove blocking risk (#10870)
- Use dynamic query stats interval and set common request timeout (#10848)
- improve internal check execution instrumentation (#10799)
Added:
- Add improved database file IO metrics and tags (#10901)
- Add statement metadata to events and metrics payload (#10881)
- Add option to disable query metrics secondary aggregates for user and database (#10975)
- Add plan handle to the plan event (#10939)
- Add the option to set a reported hostname (SQLServer) (#10688)
- Add
enforce_collection_interval_deadlineoption to set plan collection deadline (#10829) - Add current timestamp to SQLServer activity collection (#10786)
- Update SQLServer activity query to get timestamps with the timezone (#10782)
- Add active sessions monitoring support for SQLServer (#10610)
Fixed:
- Bump base package (#11067)
- Add comment to autogenerated model files (#10945)
- Fix dangling connection on unhandled exception bug (#10935)
- Improve plan lookup performance (#10828)
- Fix missing ConnectionRetryCount when DSN set (#10830)
- Fix wrong example for MultiSubnetFailover in connection_string example (#10832)
- Improve sqlserver agent query performance for DBM metrics query (#10810)
- Filter databases when using database autodiscovery (#10416)
- Update SQLServer Query Metrics Collection Query to Improve Performance (#10763)
- Add ConnectRetryCount to connection string (#10738)
Security:
- Bump lxml package (#10904)
Fixed:
Fixed:
- Fix ADO driver bugs on Windows (#10637)
Added:
Fixed:
- Handle missing DBM metrics columns on older SQL Server versions (#10594)
- Fix sqlserver resolved_hostname by handling comma correctly (#10592)
- Fix memory clerks metrics for sqlserver 2019 (#10464)
- Upgrade datadog checks base to 23.1.5 (#10468)
Changed:
- Stop sending FCI metrics when not available (#10395)
Fixed:
- Account for possible nonetype (#10257)
- DBM check should use its own connection (#10387)
- Remove duplicate names from the list of sql-server metrics to collect (#10334) Thanks pedroreys.
Fixed:
- Upgrade datadog checks base to 23.1.5 in sqlserver integration (#10468)
Fixed:
- DBM check should use its own connection (#10387)
Changed:
- Implement disable generic tags (#10290)
Added:
- Sync configs with new option and bump base requirement (#10315)
- Collect query metrics & plans for Database Monitoring (#10234)
- Disable generic tags (#10027)
Added:
- Add autodiscovered database connection service check (#9900)
Changed:
- Remove messages for integrations for OK service checks (#9888)
Added:
- Add database file metrics from sys.master_files (#9812)
Fixed:
- Capture value error (#9852)
Fixed:
- Do not throw key errors (#9460)
Added:
- Add runtime configuration validation (#8987)
Fixed:
- Fix misleading WARN message regarding adoprovider being ignored when using adodbapi connector (#9412)
Changed:
- Utilize time precision function from datadog_checks_base (#8841)
Added:
- Upgrade pywin32 on Python 3 (#8845)
Fixed:
- Fix autodiscovery tagging (#9055)
Fixed:
- Improve exception handling for database queries (#8837)
- Ensure delimited identifiers in USE statements (#8832)
- Handle availability replica metrics on earlier versions (#8830)
Fixed:
- Add availability group name tag (#8658)
- Clarify windows user and validate connection options (#8582)
- Bump minimum base package version (#8443)
Fixed:
- Fix cursor execution returning None (#8481)
Fixed:
- Avoid redundant queries (#8447)
Fixed:
- Clarify authentication in SQL Server (#8396)
Fixed:
- Handle offline databases for existence check (#8374)
- Handle overflow error for certain sql queries (#8366)
Fixed:
- Handle database specific queries for autodiscovery (#8329)
- Small refactor of consts, init and tests (#8221)
Fixed:
Added:
Fixed:
- Handle case sensitivity on database names (#8113)
- Move connection initialization outside init function (#8064)
Added:
- Add support for custom SQL queries (#8045)
- Add new database backup and fragmentation metrics for SQLServer (#7998)
Added:
- Add AlwaysOn metrics for SQLServer (#7824)
- Support additional performance metrics (#7667)
- [doc] Add encoding in log config sample (#7708)
Changed:
- SQL Server metrics refactor (#7551)
- Refactor sqlserver connection class and expand test coverage (#7510)
- Update sqlserver to Agent 6 single instance logic (#7488)
Added:
- Add Scheduler and Task Metrics for SQL Server (#5840)
Fixed:
- Fix style for the latest release of Black (#7438)
Fixed:
- Update logs config service field to optional (#7209)
Added:
- Upgrade pywin32 to 228 (#6980)
- Add default
freetdsdriver for Docker Agent (#6636) - Add log support (#6625)
Fixed:
- Fix template specs typos (#6912)
Added:
- Install
pyodbcfor MacOS and fix local test setup (#6633) - Allow optional dependency installation for all checks (#6589)
Fixed:
- Use agent 6 signature (#6447)
Fixed:
- Update deprecated imports (#6088)
Fixed:
- Streamline exception handling (#6003)
Fixed:
- Fix small capitalization error in log (#5509)
Added:
Added:
- Upgrade pywin32 to 227 (#5036)
Added:
- Upgrade pywin32 to 225 (#4563)
Added:
- Allow SQLNCLI11 provider in SQL server (#4097)
Added:
- Upgrade dependencies for Python 3.7 binary wheels (#4030)
Added:
- Adhere to code style (#3567)
Fixed:
- Don't ship
pyodbcon macOS as SQLServer integration is not shipped on macOS (#3461)
Added:
- Add custom instance tags to storedproc metrics (#3237)
Fixed:
- Use execute instead of callproc if using (py)odbc (#3236)
Added:
- Support Python 3 (#3027)
Fixed:
- Bump pyodbc for python3.7 compatibility (#2801)
Added:
- Add linux as supported OS (#2614)
Fixed:
- Additional debug logging when calling a stored procedure (#2151)
- Use raw string literals when \ is present (#2465)
Added:
- Pin pywin32 dependency (#2322)
Added:
- Support higher query granularity (#2017)
- Add ability to support (via configuration flag) the newer ADO provider (#1673)
Fixed:
- Stop leaking db password when a connection is not in the pool (#2031)
- Bump pyro4 and serpent dependencies (#2007)
- Fix for case sensitivity in the
proc_type_mappingdict. (#1860) - Add data files to the wheel package (#1727)
Added:
- support object_name metric identifiers (#1679)
Added:
- Add custom tag support for service checks.
Added:
- Allow custom connection string to connect (#1068)
Fixed:
- Allows metric collection from all instances in custom query (#959)
- Repair reporting of stats from sys.dm_os_wait_stats (#975)
Added:
- single bulk query of all of metrics, then filter locally (#573)
Added:
- Allow calling custom proc to return metrics, and improve transaction handling (#357 and #456, thanks @rlaveycalcom/rlaveycal)
Fixed:
- Fix yaml example file spacing (#342, thanks @themsquaredcom/themsquared)
Added:
- adds sqlserver integration.