Issues
- LP #1264306: Per Session innodb_flush_log_at_trx_commit poorly documentedPS-1466Resolved issue: PS-1466
- LP #1502408: When using proxy protocol, mysql must bind to an address instead of 0.0.0.0PS-3313Resolved issue: PS-3313
- ECDSA certificates fail to work with Percona ServerPS-5104
- MyRocks - Write Ahead Log (WAL) is not archived unless rocksdb_wal_ttl_seconds is set to non-zero valuePS-5425Peter Sylvester
- Audit plugin doesn't handle failed write operationsPS-6889Hrvoje Matijakovic
- mysqltest_safe_process binary is absent in .deb packagesPS-7373Resolved issue: PS-7373
- log_status has wrong binary_log_position of gtid_executedPS-7490
- MyRocks crashes with OOMPS-7494
- authentication_ldap_simple_log_status can't be 5 or more in my.cnfPS-7496
- Unable to build PerconaFT with musl on ppc64lePS-7559
- mysql crashes on index rename, or drop/add index in single alterPS-7511
- "Table_function_json..print" segmentation faultPS-7523
- Threads stuck waiting for mutex on THD::release_resourcesPS-7525Marcelo Altmann
- LOCK TABLES FOR BACKUP locking performance_schema table operationsPS-7533
- Data loss with ALL_NON_LOSSY and replication between latin1->utf8mb4PS-7546
- Assertion `innodb_trx_id == 0 || innodb_trx_id == trx_id || trx_id == 0' durin CHANGE MASTERPS-7558nitendra.bhosle
- Minor upgrade breaks P_S tables when enforce_storage_engine is setPS-7567
- Patch to force loose index scans with an optimizer hintPS-7575
- MTR tests fails to run from 5.7 build install directoryPS-7590
- system_variable_source does not report DYNAMIC when queried from sys_var updatePS-7608Yura Sorokin
- Too many columns on partitioned table makes database brokenPS-7610
- Incorrect log_bin_index status variable if log-bin-index contains periodsPS-7615
- If log_bin has two periods in the name, only a single binlog file is createdPS-7616
- PS-8.0: Assertion `rc == TYPE_OK' failedPS-7624
- [ERROR] [MY-013183] [InnoDB] Assertion failure: row0mysql.cc:2922:prebuilt->template_type == ROW_MYSQL_WHOLE_ROWPS-7628
- MultiGet-MRR (Multi-Range-Read) implementationPS-7649
- Assertion failure: fil0fil.cc:8857:err == DB_SUCCESSPS-7663Satya Bodapati
- slave server may has gaps in Executed_Gtid_Set when a special case happenPS-7673
- DDL using bulk load is very slow under long flush_listPS-7677
- The FILE_NAME OF innodb_temporary tablespace is not synced with the configured value of innodb_temp_data_file_pathPS-7680
- Improve ProcFS globbing and directory scanning efficiencyPS-7703
- compilation warning with apple clangPS-7709
- Test main.mysqld--help-notwin fails: Result content mismatchPS-7721
- Stored procedure execution fails if contains a functional indexPS-7745Daniel Fiala
- Integrate our Jenkins with GithubPS-7763Julia Vural
- [MTR] Missing end of block error with `--assert` mysqltest commandPS-7781
- RocksDB Got error 1 'NotFound: ' from ROCKSDB, Error_code: 1296; handler error No Error!PS-7791
- void cost_skip_scan(TABLE*, uint, uint, ha_rows, Cost_estimate*, ha_rows*, Item*, Opt_trace_object*): Assertion `keys_per_group >= 0'PS-7796
- Alter Table Decimal percision inserts prepends 0 after precision rather than appendPS-7803Daniel Fiala
- Configurable Index length in RocksDBPS-7812
- opt_statistics.cc:64: rec_per_key_t guess_rec_per_key(const TABLE*, const KEY*, uint): Assertion `!key->has_records_per_key(used_keyparts - 1)' failed.PS-7820
- opt_statistics.cc:115: rec_per_key_t guess_rec_per_key(const TABLE*, const KEY*, uint): Assertion `rec_per_key >= rec_per_key_all' failed.PS-7821
- Mutex deadlock around processlist, global status and global variables plus binlogsPS-7832
- Server fails to restart after crash Assertion failure: ut0ut.cc:557PS-7835
- sql_class.cc:270: virtual THD::Attachable_trx::~Attachable_trx(): Assertion `!m_thd->transaction_rollback_request' failedPS-7837
- Multi 'show binary logs' statement may block transaction commitPS-7845
- server_errno=2013 in mysql log for semisync rpl and slave_compressed_protocolPS-7850
- Assertion failure: fsp0fsp.cc:4500:!((space->flags & ((~(~0U << 1)) << (((((((0 + 1) + 4) + 1) + 4) + 1) + 1) + 1))) >> (((((((0 + 1) + 4) + 1) + 4) + 1) + 1) + 1))PS-7857
- The test `gr_startup_check_node_seed_skips_own_address` fails on JenkinsPS-7863
- MyRocks encryption - implementationPS-7887Kamil Holubicki
50 of 280
LP #1264306: Per Session innodb_flush_log_at_trx_commit poorly documented
Done
General
Escalation
General
Escalation
Description
Environment
None
Smart Checklist
Details
Assignee
UnassignedUnassignedReporter
lpjirasynclpjirasync(Deactivated)Labels
Needs Review
YesTime tracking
7h 40m loggedComponents
Fix versions
Priority
High
Details
Details
Assignee
Unassigned
UnassignedReporter
lpjirasync
lpjirasync(Deactivated)Labels
Needs Review
Yes
Time tracking
7h 40m logged
Components
Fix versions
Priority
Smart Checklist
Smart Checklist
Smart Checklist
Created January 21, 2018 at 4:00 PM
Updated March 6, 2024 at 3:09 PM
Resolved March 24, 2022 at 12:08 PM
Activity
Show:
patrick.birchJanuary 17, 2020 at 4:01 PM
Merged 5.6 version
Merged 5.7 version
lpjirasyncJanuary 21, 2018 at 4:01 PM
**Comment from Launchpad by: Jaime Crespo on: 27-12-2013 09:39:24
For future reference, enabling a feature by setting it from 1 to 0 is anti-intuitive. I understand the original requirement of maintaining backwards compatibility, but setting innodb_use_session_flush_log_at_trx_commit to 1 would have been more intuitive.
lpjirasyncJanuary 21, 2018 at 4:01 PM
**Comment from Launchpad by: Laurynas Biveinis on: 27-12-2013 04:23:03
I agree that innodb_use_global_flush_log_at_trx_commit docs should be improved. The 5.1 innodb_flush_log_at_trx_commit_session docs are IMHO clearer but still targeting the bug to 5.1 too for a review (e.g. is innodb_flush_log_at_trx_commit_session really a global and not session variable?)
**Reported in Launchpad by Jaime Crespo last update 01-08-2017 06:48:14
This and the same section on 5.5 is poorly written and has some documentation bugs:
http://www.percona.com/doc/percona-server/5.6/scalability/innodb_io.html#innodb_use_global_flush_log_at_trx_commit
In summary:
The feature doc is difficult to understand, and overuses "GLOBAL" when it means "session as if it was set globally"
I think it has some erroneous information: wrong quoted example and the session variable can always be set, but (hopefully) it will have no effect
The feature is not documented in any other place, and it requires further discussion than simple parameters settings
I have written an alternative version (please check validity) to at least change the current information:
---------------
This variable is used to enable or disable the effect of the per session value of innodb_flush_log_at_trx_commit.
If the global variable innodb_use_global_flush_log_at_trx_commit is set to 1 (True), the session will use always use the current global value of innodb_flush_log_at_trx_commit. This is the default value.
Under this mode, changing the session variable for innodb_flush_log_at_trx_commit will have no effect for upstream compatibility.
If the global variable innodb_use_global_flush_log_at_trx_commit is set to 0 (False), the user can modify the MySQL commit durability per session, using:
SET SESSION innodb_flush_log_at_trx_commit=N
This change will only affect the transactions in the session where it has changed, while the rest will keep the global innodb_flush_log_at_trx_commit value unless similarly changed.