- About TiDB
- Quick Start
- Deploy
- Software and Hardware Requirements
- Environment Configuration Checklist
- Topology Patterns
- Install and Start
- Verify Cluster Status
- Benchmarks Methods
- Migrate
- Maintain
- Monitor and Alert
- Troubleshoot
- TiDB Troubleshooting Map
- Identify Slow Queries
- Analyze Slow Queries
- SQL Diagnostics
- Identify Expensive Queries
- Statement Summary Tables
- Troubleshoot Hotspot Issues
- Troubleshoot Increased Read and Write Latency
- Troubleshoot Cluster Setup
- Troubleshoot High Disk I/O Usage
- Troubleshoot Lock Conflicts
- Troubleshoot TiFlash
- Troubleshoot Write Conflicts in Optimistic Transactions
- Performance Tuning
- System Tuning
- Software Tuning
- SQL Tuning
- Overview
- Understanding the Query Execution Plan
- SQL Optimization Process
- Overview
- Logic Optimization
- Physical Optimization
- Prepare Execution Plan Cache
- Control Execution Plans
- Tutorials
- TiDB Tools
- Overview
- Use Cases
- Download
- TiUP
- TiDB Operator
- Backup & Restore (BR)
- TiDB Binlog
- TiDB Lightning
- TiDB Data Migration
- TiCDC
- Dumpling
- sync-diff-inspector
- Loader
- Mydumper
- Syncer
- TiSpark
- Reference
- Cluster Architecture
- Key Monitoring Metrics
- Secure
- Privileges
- SQL
- SQL Language Structure and Syntax
- SQL Statements
ADD COLUMN
ADD INDEX
ADMIN
ADMIN CANCEL DDL
ADMIN CHECKSUM TABLE
ADMIN CHECK [TABLE|INDEX]
ADMIN SHOW DDL [JOBS|QUERIES]
ALTER DATABASE
ALTER INSTANCE
ALTER TABLE
ALTER USER
ANALYZE TABLE
BACKUP
BEGIN
CHANGE COLUMN
COMMIT
CHANGE DRAINER
CHANGE PUMP
CREATE [GLOBAL|SESSION] BINDING
CREATE DATABASE
CREATE INDEX
CREATE ROLE
CREATE SEQUENCE
CREATE TABLE LIKE
CREATE TABLE
CREATE USER
CREATE VIEW
DEALLOCATE
DELETE
DESC
DESCRIBE
DO
DROP [GLOBAL|SESSION] BINDING
DROP COLUMN
DROP DATABASE
DROP INDEX
DROP ROLE
DROP SEQUENCE
DROP STATS
DROP TABLE
DROP USER
DROP VIEW
EXECUTE
EXPLAIN ANALYZE
EXPLAIN
FLASHBACK TABLE
FLUSH PRIVILEGES
FLUSH STATUS
FLUSH TABLES
GRANT <privileges>
GRANT <role>
INSERT
KILL [TIDB]
LOAD DATA
LOAD STATS
MODIFY COLUMN
PREPARE
RECOVER TABLE
RENAME INDEX
RENAME TABLE
REPLACE
RESTORE
REVOKE <privileges>
REVOKE <role>
ROLLBACK
SELECT
SET DEFAULT ROLE
SET [NAMES|CHARACTER SET]
SET PASSWORD
SET ROLE
SET TRANSACTION
SET [GLOBAL|SESSION] <variable>
SHOW ANALYZE STATUS
SHOW [BACKUPS|RESTORES]
SHOW [GLOBAL|SESSION] BINDINGS
SHOW BUILTINS
SHOW CHARACTER SET
SHOW COLLATION
SHOW [FULL] COLUMNS FROM
SHOW CONFIG
SHOW CREATE SEQUENCE
SHOW CREATE TABLE
SHOW CREATE USER
SHOW DATABASES
SHOW DRAINER STATUS
SHOW ENGINES
SHOW ERRORS
SHOW [FULL] FIELDS FROM
SHOW GRANTS
SHOW INDEX [FROM|IN]
SHOW INDEXES [FROM|IN]
SHOW KEYS [FROM|IN]
SHOW MASTER STATUS
SHOW PLUGINS
SHOW PRIVILEGES
SHOW [FULL] PROCESSSLIST
SHOW PROFILES
SHOW PUMP STATUS
SHOW SCHEMAS
SHOW STATS_HEALTHY
SHOW STATS_HISTOGRAMS
SHOW STATS_META
SHOW STATUS
SHOW TABLE NEXT_ROW_ID
SHOW TABLE REGIONS
SHOW TABLE STATUS
SHOW [FULL] TABLES
SHOW [GLOBAL|SESSION] VARIABLES
SHOW WARNINGS
SHUTDOWN
SPLIT REGION
START TRANSACTION
TRACE
TRUNCATE
UPDATE
USE
- Data Types
- Functions and Operators
- Overview
- Type Conversion in Expression Evaluation
- Operators
- Control Flow Functions
- String Functions
- Numeric Functions and Operators
- Date and Time Functions
- Bit Functions and Operators
- Cast Functions and Operators
- Encryption and Compression Functions
- Information Functions
- JSON Functions
- Aggregate (GROUP BY) Functions
- Window Functions
- Miscellaneous Functions
- Precision Math
- List of Expressions for Pushdown
- Constraints
- Generated Columns
- SQL Mode
- Transactions
- Garbage Collection (GC)
- Views
- Partitioning
- Character Set and Collation
- System Tables
mysql
- INFORMATION_SCHEMA
- Overview
ANALYZE_STATUS
CHARACTER_SETS
CLUSTER_CONFIG
CLUSTER_HARDWARE
CLUSTER_INFO
CLUSTER_LOAD
CLUSTER_LOG
CLUSTER_SYSTEMINFO
COLLATIONS
COLLATION_CHARACTER_SET_APPLICABILITY
COLUMNS
DDL_JOBS
ENGINES
INSPECTION_RESULT
INSPECTION_RULES
INSPECTION_SUMMARY
KEY_COLUMN_USAGE
METRICS_SUMMARY
METRICS_TABLES
PARTITIONS
PROCESSLIST
SCHEMATA
SEQUENCES
SESSION_VARIABLES
SLOW_QUERY
STATISTICS
TABLES
TABLE_CONSTRAINTS
TABLE_STORAGE_STATS
TIDB_HOT_REGIONS
TIDB_INDEXES
TIDB_SERVERS_INFO
TIFLASH_REPLICA
TIKV_REGION_PEERS
TIKV_REGION_STATUS
TIKV_STORE_STATUS
USER_PRIVILEGES
VIEWS
METRICS_SCHEMA
- UI
- TiDB Dashboard
- Overview
- Maintain
- Access
- Overview Page
- Cluster Info Page
- Key Visualizer Page
- Metrics Relation Graph
- SQL Statements Analysis
- Slow Queries Page
- Cluster Diagnostics
- Search Logs Page
- Profile Instances Page
- Session Management and Configuration
- FAQ
- CLI
- Command Line Flags
- Configuration File Parameters
- System Variables
- Storage Engines
- Telemetry
- Errors Codes
- Table Filter
- Schedule Replicas by Topology Labels
- FAQs
- Release Notes
- All Releases
- v4.0
- v3.1
- v3.0
- v2.1
- v2.0
- v1.0
- Glossary
Troubleshoot Write Conflicts in Optimistic Transactions
This document introduces the reason of and solutions to write conflicts in optimistic transactions.
Before TiDB v3.0.8, TiDB uses the optimistic transaction model by default. In this model, TiDB does not check conflicts during transaction execution. Instead, while the transaction is finally committed, the two-phase commit (2PC) is triggered and TiDB checks write conflicts. If a write conflict exists and the auto-retry mechanism is enabled, then TiDB retries the transaction within limited times. If the retry succeeds or has reached the upper limit on retry times, TiDB returns the result of transaction execution to the client. Therefore, if a lot of write conflicts exist in the TiDB cluster, the duration can be longer.
The reason of write conflicts
TiDB implements its transactions by using the Percolator transaction model. percolator
is generally an implementation of 2PC. For the detailed 2PC process, see TiDB Optimistic Transaction Model.
After the client sends a COMMIT
request to TiDB, TiDB starts the 2PC process:
- TiDB chooses one key from all keys in the transaction as the primary key of the transaction.
- TiDB sends the
prewrite
request to all the TiKV Regions involved in this commit. TiKV judges whether all keys can preview successfully. - TiDB receives the result that all
prewrite
requests are successful. - TiDB gets the
commit_ts
from PD. - TiDB sends the
commit
request to the TiKV Region that contains the primary key of the transaction. After TiKV receives thecommit
request, it checks the validity of the data and clears the locks left in theprewrite
stage. - After the
commit
request returns successfully, TiDB returns success to the client.
The write conflict occurs in the prewrite
stage. When the transaction finds that another transaction is writing the current key (data.commit_ts
> txn.start_ts
), a write conflict occurs.
Detect write conflicts
In the TiDB Grafana panel, check the following monitoring metrics under KV Errors:
KV Backoff OPS indicates the count of error messages per second returned by TiKV.
The
txnlock
metric indicates the write-write conflict. ThetxnLockFast
metric indicates the read-write conflict.Lock Resolve OPS indicates the count of items related to transaction conflicts per second:
not_expired
indicates the TTL of the lock was not expired. The conflict transaction cannot resolve locks until the TTL is expired.wait_expired
indicates that the transaction needs to wait the lock to expire.expired
indicates the TTL of the lock was expired. Then the conflict transaction can resolve this lock.
KV Retry Duration indicates the duration of re-sends the KV request:
You can also use [kv:9007]Write conflict
as the key word to search in the TiDB log. The key word also indicates the write conflict exists in the cluster.
Resolve write conflicts
If many write conflicts exist in the cluster, it is recommended to find out the write conflict key and the reason, and then try to change the application logic to avoid write conflicts. When the write conflict exists in the cluster, you can see the log similar to the following one in the TiDB log file:
[2020/05/12 15:17:01.568 +08:00] [WARN] [session.go:446] ["commit failed"] [conn=3] ["finished txn"="Txn{state=invalid}"] [error="[kv:9007]Write conflict, txnStartTS=416617006551793665, conflictStartTS=416617018650001409, conflictCommitTS=416617023093080065, key={tableID=47, indexID=1, indexValues={string, }} primary={tableID=47, indexID=1, indexValues={string, }} [try again later]"]
The explanation of the log above is as follows:
[kv:9007]Write conflict
: indicates the write-write conflict.txnStartTS=416617006551793665
:indicates thestart_ts
of the current transaction. You can use thepd-ctl
tool to convertstart_ts
to physical time.conflictStartTS=416617018650001409
: indicates thestart_ts
of the write conflict transaction.conflictCommitTS=416617023093080065
: indicates thecommit_ts
of the write conflict transaction.key={tableID=47, indexID=1, indexValues={string, }}
:indicates the write conflict key.tableID
indicates the ID of the write conflict table.indexID
indicates the ID of write conflict index. If the write conflict key is a record key, the log printshandle=x
, indicating which record(row) has a conflict.indexValues
indicates the value of the index that has a conflict.primary={tableID=47, indexID=1, indexValues={string, }}
: indicates the primary key information of the current transaction.
You can use the pd-ctl
tool to convert the timestamp to readable time:
./pd-ctl -u https://127.0.0.1:2379 tso {TIMESTAMP}
You can use tableID
to find the name of the related table:
curl http://{TiDBIP}:10080/db-table/{tableID}
You can use indexID
and the table name to find the name of the related index:
SELECT * FROM INFORMATION_SCHEMA.TIDB_INDEXES WHERE TABLE_SCHEMA='{table_name}' AND TABLE_NAME='{table_name}' AND INDEX_ID={indexID};
In addition, in TiDB v3.0.8 and later versions, the pessimistic transaction becomes the default model. The pessimistic transaction model can avoid write conflicts during the transaction prewrite stage, so you do not need to modify the application any more. In the pessimistic transaction mode, each DML statement writes a pessimistic lock to the related keys during execution. This pessimistic lock can prevent other transactions from modifying the same keys, thus ensuring no write conflicts exist in the prewrite
stage of the transaction 2PC.