What's about to use more general approach and collect statistics into special read-only caches ?
It gives more advantages like using data from such caches for other (user) queries, track statistics without coding, behavior is closer to RDBMS, no new API. p.s. For instance take a look on http://www.postgresql.org/docs/current/static/pgstatstatements.html On Wed, Apr 27, 2016 at 9:37 PM, Dmitriy Setrakyan <[email protected]> wrote: > Ideally, it would be good to provide average time for every SQL statement. > Do we track it anywhere? > > On Wed, Apr 27, 2016 at 7:01 AM, Sergi Vladykin <[email protected]> > wrote: > > > I'm not sure I understand, what kind of statistics do you need. Probably > we > > could fire events for slow queries (I don't think it is a good idea to > > generate an even for any query due to performance reasons) and then > you'll > > be able to add any listeners you want. > > > > Sergi > > > > 2016-04-27 3:39 GMT+03:00 Alexey Kuznetsov <[email protected]>: > > > > > I'm going to implement support for SQL statements monitoring and > > statistics > > > in Web Console. > > > > > > For this feature we need API that will allow to collect info about what > > > statements were executed in cluster and their metrics. > > > > > > @Sergi Vladykin, as you one of the most experts for SQL subsystem, > could > > > you tell me, > > > where and how to implement such API? > > > > > > Thanks! > > > > > > -- > > > Alexey Kuznetsov > > > GridGain Systems > > > www.gridgain.com > > > > > > -- Sergey Kozlov GridGain Systems www.gridgain.com
