my speak about MySQL bugs at FOSDEM 2019 MySQL, MariaDB and Friends Devroon are all set, support consumers chose not to break anything terribly on weekend, so I have some spare time for blogging. As normal, when I do not have any much better concept or beneficial current reality experience to share I blog about MySQL bugs . Today I ‘d like to continue my evaluation of intriguing MySQL bug reports included by Community members in December, 2018. I’ll examine them beginning with the earliest: Bug # 93701 – “Assertion’ maybe_null ‘stopped working|Item_func_concat:: val_str( String *)”. It’s debug assertion, not a huge offer (S6 bug), so why should anybody appreciate this report from Ramesh Sivaraman ? Since last time exact same assertion failure wound up as major adequate Bug # 83115 repaired in 8.0.11. Bug # 93708 – “Page Cleaner will sleep for long period of time if clock modifications”. In this bug report from Marcelo Altmann I was primarily impressed by arguing begun by Oracle engineer on bug seriousness (paperwork demand, function demand vs genuine bug). IMHO it was a wild-goose chase. Bug # 93728 – “mysqld crash after utilizing change table with SPATIAL secret”. Good finding by Rui Xu . MariaDB 10.3.7 is likewise impacted. A minimum of in the mistake log I likewise see this mistake message for each SPATIAL index of the table:2019 -01 -20 19:26:15 9 [MISTAKE] InnoDB: Record in index ‘idx2’ of table ‘test’.’ tab’ was not discovered on upgrade: TUPLE (info_bits= 0, 2 fields): [32] $@ $@ $@ $@( 0x0000000000000400000000000000040000000000000004000000000000000400), [4] (0x00000001) at: COMPACT RECORD( info_bits= 0, 1 fields): [8] infimum ( 0x090E06090D050D00) Bug # 93734 – “MySQL 8.0 is 36 times slower than MySQL 5.7”. I like whatever about this bug reported by Vadim Tkachenko , CTO of Percona. I like summary that sounds cool, missing out on precise MySQL 8.0.x variation (however I presume 8.0.13), effect of innodb_flush_log_at_trx_commit, recommended repair and the reality that the bug is still “Open” and no one from Oracle paid any attention to the report from CTO of among crucial Oracle partners and advocates of MySQL 8. Terrific! Bug # 93737 – “mysqlbinlog mermory utilized grows unstoped”. I had actually not attempted to replicate it (and no one had it appears, as the bug is still “Open”), however this report by Chandler Bong sounded fascinating. Let’s see what it might wind up with … Bug # 93746 – “MySQL Crashes with deadlock at Mutex DICT_SYS developed dict0dict.cc:1172”. Long semaphore waits including DICT_SYS mutex and resulting in guard dog thread eliminating MySQL prevail. Bug processing is still in development, however I signed up for this bug by Anton Ravich primarily to follow another bug he describes as a potentially associated, Bug # 80919 . That bug is stated a replicate of some bug in Oracle’s internal bugs database, without bug number discussed. This is not the very first time I see such am creative method to procedure neighborhood bugs (that I think about undesirable and completely incorrect). Bug # 93760 – “InnoDB got assert failure while finding out area id at start-up”. Fungo Wang discovered this crash while running MTR test case innodb.innodb _ redo_debug_1 with debug binaries. Now I question if Oracle QA runs MTR tests with debug binaries at all and do they care to inspect failures (or depend on Percona for this activity)? Any remarks? Bug # 93761 – “enhance table trigger the servant MTS deadlock!”. Sounds major and fascinating, however no one cared to inspect this bug report. It’s “Open” without remarks for 3 weeks currently. Bug # 93767 – “INSTALL COMPONENT stops working – managed segfault throughout setup absolutely nothing in log”. I am amazed that someone beyond Oracle currently utilizes brand-new MySQL 8.0 server elements function. My previous associate, Justin Swanhart , attempted however stopped working up until now. Let’s see how this report might wind up. In the meantime Oracle engineer mentioned that Justin attempted to do something this structure was not developed for, see good last remark there. Bug # 93777 – “the Originator of mysql.event is not remedy”. Take notice of this bug discovered by Dennis Gao if you utilize occasions in duplication environment. Bug # 93779 – “dml in table with trigger to other Table map in binlog result in servant sql stopped”. Deal with this bug report from Mohamed Atef is still in development. Bug reported cared to include a great deal of information in current remarks, so I want to see this bug processed quickly.I’ve likewise consisted of a number of bugs reported in January, 2019 in the list above, however I hope you do not mind. Today I ‘d like to keep in mind those warm days in Venice in September and those bothersome MySQL bugs reported this winter season …To sum up:Some bug reports remain “Open” for weeks with no great factor just recently.The disgraceful practice of describing bugs in Oracle’s internal bugs database without providing a number (that we see in release notes when the bug is repaired and can try to find in dedicate messages at GitHub) ought to be stopped! Next time I’ll blame engineer(s) who this by name, here and in social networks. Inspect how this is effectively done by Shane Bester.I question if Oracle QA runs MTR tests cases on debug constructs on a regula basis and checks the outcomes, or this is now a task for Percona mainly?I need to compose a different article about bugs associated with SPATIAL indexes execution in InnoDB. There need to be dragons …P.S. Don’t forget to click names of bug press reporters to see the whole list of still active MySQL bug reports each of them developed.

Read more: mysqlentomologist.blogspot.com

Leave a comment

Your email address will not be published. Required fields are marked *