You tripled the amount of memory assigned to the cache, but your Qcache_lowmem_prunes value went up 10x and Qcache_not_cached went up 6x. I would say that you have enough queries that are unique, that most of your queries are getting cached then flushed before the same query comes up again.

The number of queries being cached went up 13x, so you are caching a lot more queries, but it's not doing all that much good. You're getting diminishing returns as you increase memory. The query cache works best when you have the same exact queries being run over and over. You know your code, so you need to think how many unique queries you have. Perhaps you're getting close to the number of unique queries you run and increasing your cache just a bit more will suddenly send your hit rate way up, perhaps not. Test in big increments. If you suddenly get a big boost in hit ratio, work your way back down. But it may be that the query cache just can't be used optimally with your system.


On Sep 15, 2005, at 1:10 PM, Anil wrote:





  _____

From: Anil [mailto:[EMAIL PROTECTED]
Sent: Thursday, September 15, 2005 10:36 PM
To: 'mysql@lists.mysql.com'
Subject: QUERY CACHE TUNING QUESTION



Hi list,



We are using mysql 4.0.24 on RHEL 3.0 on dell 2650 machine with 2 gb RAM.
We tried to fine tune query cache . please find below the analysis







 PARAMETER                             old                  modified

 ==========               ==========     ================



query_cache_size                 10 MB                   32 MB



Qcache_queries_in_cache      187                       2444



Qcache_inserts                     1226682                13560221



Qcache_hits                          1510151                 7380756



Qcache_lowmem_prunes         850128                8096499



Qcache_not_cached                2725692              16361318



Qcache_free_blocks                   65                    12



Qcache_free_memory                9.26 MB           28.79 MB







qCACHE HIT RATE                    27.64%          19.78%







QCACHE HIT RATE IS CALUCULATED WITH FORMULA = (Qcache_hits/ (Qcache_inserts
+Qcache_not_cached+Qcache_hits))  * 100



QCACHE HIT RATE IS CALUCULATED WITH FORMULA = (Qcache_hits/ (Qcache_inserts
+Qcache_not_cached+Qcache_hits))  * 100



As per the above analysis we observed that qcache hit rate came down after
increasing query_cache_size from 10 MB to 32 MB and  qcache_hits got
increased but the remaining parameters Qcache_inserts, Qcache_not_cached,
Qcache_lowmem_prunes also increased because of that hit rate came down
drastically. We didn't understand the reason behind hit . Is there any
recommended procedure to fine tune query cache . awaiting early reply from
list.





Thanks in advance

Anil

DBA























--
Brent Baisley
Systems Architect
Landover Associates, Inc.
Search & Advisory Services for Advanced Technology Environments
p: 212.759.6400/800.759.0577



--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:    http://lists.mysql.com/[EMAIL PROTECTED]

Reply via email to