SQL SERVER – How to get historical deadlock Information from System Health Extended Events?

Let me start off by asking a simple question. How many of you have seen this error about historical deadlock information earlier?

Msg 1205, Level 13, State 45, Line 4
Transaction (Process ID 52) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.

I am sure most of us who work with SQL Server would say “Yes!”

Now, next question is – what would you do if you see such error? Few DBAs would say; we need to enable trace flag 1222 and wait for the next occurrence so that information is recorded in the SQL Server ERRORLOG file.

Let me share a piece of advice. There is a session called as System_Health which is created in SQL Server and captured a lot of extended events by default.

Here is the script of the session.

CREATE EVENT SESSION [system_health] ON SERVER 
ADD EVENT sqlclr.clr_allocation_failure(
    ACTION(package0.callstack,sqlserver.session_id)),
ADD EVENT sqlclr.clr_virtual_alloc_failure(
    ACTION(package0.callstack,sqlserver.session_id)),
ADD EVENT sqlos.memory_broker_ring_buffer_recorded,
ADD EVENT sqlos.memory_node_oom_ring_buffer_recorded(
    ACTION(package0.callstack,sqlserver.session_id,sqlserver.sql_text,sqlserver.tsql_stack)),
ADD EVENT sqlos.process_killed(
    ACTION(package0.callstack,sqlserver.client_app_name,sqlserver.client_hostname,sqlserver.client_pid,sqlserver.query_hash,sqlserver.session_id,sqlserver.session_nt_username)),
ADD EVENT sqlos.scheduler_monitor_deadlock_ring_buffer_recorded,
ADD EVENT sqlos.scheduler_monitor_non_yielding_iocp_ring_buffer_recorded,
ADD EVENT sqlos.scheduler_monitor_non_yielding_ring_buffer_recorded,
ADD EVENT sqlos.scheduler_monitor_non_yielding_rm_ring_buffer_recorded,
ADD EVENT sqlos.scheduler_monitor_stalled_dispatcher_ring_buffer_recorded,
ADD EVENT sqlos.scheduler_monitor_system_health_ring_buffer_recorded,
ADD EVENT sqlos.wait_info(
    ACTION(package0.callstack,sqlserver.session_id,sqlserver.sql_text)
    WHERE ([duration]>(15000) AND ([wait_type]>=N'LATCH_NL' AND ([wait_type]>=N'PAGELATCH_NL' AND [wait_type]<=N'PAGELATCH_DT' OR [wait_type]<=N'LATCH_DT' OR [wait_type]>=N'PAGEIOLATCH_NL' AND [wait_type]<=N'PAGEIOLATCH_DT' OR [wait_type]>=N'IO_COMPLETION' AND [wait_type]<=N'NETWORK_IO' OR [wait_type]=N'RESOURCE_SEMAPHORE' OR [wait_type]=N'SOS_WORKER' OR [wait_type]>=N'FCB_REPLICA_WRITE' AND [wait_type]<=N'WRITELOG' OR [wait_type]=N'CMEMTHREAD' OR [wait_type]=N'TRACEWRITE' OR [wait_type]=N'RESOURCE_SEMAPHORE_MUTEX') OR [duration]>(30000) AND [wait_type]<=N'LCK_M_RX_X'))), ADD EVENT sqlos.wait_info_external( ACTION(package0.callstack,sqlserver.session_id,sqlserver.sql_text) WHERE ([duration]>(5000) AND ([wait_type]>=N'PREEMPTIVE_OS_GENERICOPS' AND [wait_type]<=N'PREEMPTIVE_OS_ENCRYPTMESSAGE' OR [wait_type]>=N'PREEMPTIVE_OS_INITIALIZESECURITYCONTEXT' AND [wait_type]<=N'PREEMPTIVE_OS_QUERYSECURITYCONTEXTTOKEN' OR [wait_type]>=N'PREEMPTIVE_OS_AUTHZGETINFORMATIONFROMCONTEXT' AND [wait_type]<=N'PREEMPTIVE_OS_REVERTTOSELF' OR [wait_type]>=N'PREEMPTIVE_OS_CRYPTACQUIRECONTEXT' AND [wait_type]<=N'PREEMPTIVE_OS_DEVICEOPS' OR [wait_type]>=N'PREEMPTIVE_OS_NETGROUPGETUSERS' AND [wait_type]<=N'PREEMPTIVE_OS_NETUSERMODALSGET' OR [wait_type]>=N'PREEMPTIVE_OS_NETVALIDATEPASSWORDPOLICYFREE' AND [wait_type]<=N'PREEMPTIVE_OS_DOMAINSERVICESOPS' OR [wait_type]=N'PREEMPTIVE_OS_VERIFYSIGNATURE' OR [duration]>(45000) AND ([wait_type]>=N'PREEMPTIVE_OS_SETNAMEDSECURITYINFO' AND [wait_type]<=N'PREEMPTIVE_CLUSAPI_CLUSTERRESOURCECONTROL' OR [wait_type]>=N'PREEMPTIVE_OS_RSFXDEVICEOPS' AND [wait_type]<=N'PREEMPTIVE_OS_DSGETDCNAME' OR [wait_type]>=N'PREEMPTIVE_OS_DTCOPS' AND [wait_type]<=N'PREEMPTIVE_DTC_ABORT' OR [wait_type]>=N'PREEMPTIVE_OS_CLOSEHANDLE' AND [wait_type]<=N'PREEMPTIVE_OS_FINDFILE' OR [wait_type]>=N'PREEMPTIVE_OS_GETCOMPRESSEDFILESIZE' AND [wait_type]<=N'PREEMPTIVE_ODBCOPS' OR [wait_type]>=N'PREEMPTIVE_OS_DISCONNECTNAMEDPIPE' AND [wait_type]<=N'PREEMPTIVE_CLOSEBACKUPMEDIA' OR [wait_type]=N'PREEMPTIVE_OS_AUTHENTICATIONOPS' OR [wait_type]=N'PREEMPTIVE_OS_FREECREDENTIALSHANDLE' OR [wait_type]=N'PREEMPTIVE_OS_AUTHORIZATIONOPS' OR [wait_type]=N'PREEMPTIVE_COM_COCREATEINSTANCE' OR [wait_type]=N'PREEMPTIVE_OS_NETVALIDATEPASSWORDPOLICY' OR [wait_type]=N'PREEMPTIVE_VSS_CREATESNAPSHOT')))), ADD EVENT sqlserver.connectivity_ring_buffer_recorded(SET collect_call_stack=(1)), ADD EVENT sqlserver.error_reported( ACTION(package0.callstack,sqlserver.database_id,sqlserver.session_id,sqlserver.sql_text,sqlserver.tsql_stack) WHERE ([severity]>=(20) OR ([error_number]=(17803) OR [error_number]=(701) OR [error_number]=(802) OR [error_number]=(8645) OR [error_number]=(8651) OR [error_number]=(8657) OR [error_number]=(8902) OR [error_number]=(41354) OR [error_number]=(41355) OR [error_number]=(41367) OR [error_number]=(41384) OR [error_number]=(41336) OR [error_number]=(41309) OR [error_number]=(41312) OR [error_number]=(41313)))),
ADD EVENT sqlserver.security_error_ring_buffer_recorded(SET collect_call_stack=(1)),
ADD EVENT sqlserver.sp_server_diagnostics_component_result(SET collect_data=(1)
    WHERE ([sqlserver].[is_system]=(1) AND [component]<>(4))),
ADD EVENT sqlserver.sql_exit_invoked(
    ACTION(package0.callstack,sqlserver.client_app_name,sqlserver.client_hostname,sqlserver.client_pid,sqlserver.query_hash,sqlserver.session_id,sqlserver.session_nt_username)),
ADD EVENT sqlserver.xml_deadlock_report
ADD TARGET package0.event_file(SET filename=N'system_health.xel',max_file_size=(5),max_rollover_files=(4)),
ADD TARGET package0.ring_buffer(SET max_events_limit=(5000),max_memory=(4096))
WITH (MAX_MEMORY=4096 KB,EVENT_RETENTION_MODE=ALLOW_SINGLE_EVENT_LOSS,MAX_DISPATCH_LATENCY=120 SECONDS,MAX_EVENT_SIZE=0 KB,MEMORY_PARTITION_MODE=NONE,TRACK_CAUSALITY=OFF,STARTUP_STATE=ON)
GO

There is no need to worry about a lot of information there. The information which I want to highlight is below.

ADD EVENT sqlserver.xml_deadlock_report

Above means that by default SQL Server should be capturing XML deadlock graph. To check this I have done little experiment.

Create Database and Objects

First create a database and two tables.

CREATE DATABASE DeadlockDemo
GO
USE DeadlockDemo
GO
CREATE TABLE MyT1 (i INT)
GO
CREATE TABLE MyT2 (i INT)
GO
INSERT INTO MyT1
VALUES (1)
GO
INSERT INTO MyT2
VALUES (1)
GO

Reproduce Deadlock Error

We need to connections to reproduce deadlock.

Once above steps are complete, once of the connection would get deadlock error “Transaction (Process ID 52) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.”

Look at the deadlock graph from System health session:

Here is the query which we can use to read deadlock graph which is captured in system health session.

SELECT XEvent.query('(event/data/value/deadlock)[1]') AS DeadlockGraph
FROM (
	SELECT XEvent.query('.') AS XEvent
	FROM (
		SELECT CAST(target_data AS XML) AS TargetData
		FROM sys.dm_xe_session_targets st
		INNER JOIN sys.dm_xe_sessions s ON s.address = st.event_session_address
		WHERE s.NAME = 'system_health'
			AND st.target_name = 'ring_buffer'
		) AS Data
CROSS APPLY TargetData.nodes('RingBufferTarget/event[@name="xml_deadlock_report"]') AS XEventData(XEvent)
) AS source;

And we will get same graph which we get via trace flag 1222 in the error log.

Do above experiment and share your experience via the comments.

Reference: Pinal Dave (http://blog.sqlauthority.com)

First appeared on SQL SERVER – How to get historical deadlock Information from System Health Extended Events?