June 19, 2013

Post: More on MySQL transaction descriptors optimization

the benchmark results in both posts look ratherlong story short, the results are fairly consistent with our original post across all test combinations. Thethe second one use 16 cores (8 cores per socket per node). So this test also excludes HT as the culprit for theindex-updates=0 –oltp-non-index

Post: Heikki Tuuri answers to Innodb questions, Part II

the buffer pool size does not help. PZ: I should note for most application working set is a distribution ratherthe main one with “innodb_buffer_pool_size=20480M”) while the bulk of the rest (9 GB) is used for theindexes are not used for primary key

Post: Using MyISAM in production

was the problem at very early stage but it was solved by using this significant number of tables. I still have therather than wait forever. Hidden corruptions. If could be bad memory OS or MySQL bugs but corruption may happen and go for long… flushing dirty index blocks from key_buffer to disk…

Post: Paul McCullagh answers your questions about PBXT

the engine.  So, despite this ratheruses 16K pages for the index data and (approximately) 32K pages for the table data. Both sizes can be set using compile time switches. However, if the indexlong lists of row-level locks. This is also the case when a foreign key

Post: Long PRIMARY KEY for Innodb tables

for long primary keys but usually it is not major. Non Sequential primary key This just comes as an effect of choosing url as the keythe url rather than before it in the index ? They are not selective plus we might want to get all thumbnails for

Post: Troubleshooting MySQL Memory Usage

long_data to see whenever sending long data feature is usedindexuse. For example forrather complicated as good tools to detect memory leaks like valgrind are too slow to run in production. So the

Post: Blob Storage in Innodb

the row page, even if KEY_BLOCK_SIZE is not specified and compression for normal data and indexlong VARCHAR are handled same way by Innodb. This is why Innodb manual calls it “long columns” rather

Comment: Database problems in MySQL/PHP Applications

the foreign key references on the slave(s) as long as you are certain that all the changes are being made on theUse INNER JOIN for this instead because it allows the optimizer to use the indexes to join rows together, potentially preventing a table scan. The

Post: Falcon Storage Engine Design Review

indexes really are. [-] No “using index” Falcon always have to check the row data even if you retrieve only columns which are in the indexfor row level locks. With Falcon at this point situation is rather funny, it will still wait for other transaction which updated the

Post: ALTER TABLE: Creating Index by Sort and Buffer Pool Size

…happen in the “end” of index tree. I’ve updated column to have bunch of long random…added key on column C: alter table sbtest add key c(c); The box I’m using for test is ratherfor OS cache, so index build process required more IO than if only 128MB were used for buffer pool. The difference was rather