FWD » Core Development » Database

03/28/2022

06:53 AM Bug #3896: improve performance of the _lock metadata implementation

Sorry, forgot to add new classes to the .diff file. Igor Skornyakov

06:02 AM Bug #3896: improve performance of the _lock metadata implementation

Finished.
Please review.
Thank you. Igor Skornyakov

03/27/2022

03:21 PM Bug #3896: improve performance of the _lock metadata implementation

Added JMX instrumentation and cleanup logic to @LockTableUpdater@.
However, I have a strange problem with finding pe... Igor Skornyakov

03/25/2022

07:34 AM Bug #3896: improve performance of the _lock metadata implementation

Please note that the @_Lock@ VST cleanup in a separate thread should be a rare event. It will happen only if the lock... Igor Skornyakov

07:11 AM Bug #3896: improve performance of the _lock metadata implementation

Greg Shah wrote:
> I prefer to avoid executors or Java thread pools. It adds hidden logic to manage threads which m... Igor Skornyakov

06:49 AM Bug #3896: improve performance of the _lock metadata implementation

> I'm planning to use a @SingleThreadExecutor@ shared by all instances of the @LockTableUpdater@ for the background @... Greg Shah

05:07 AM Bug #3896: improve performance of the _lock metadata implementation

After fixing an obvious mistake with adding @final@ to the @Empty@ class declaration the smoke test of the large cust... Igor Skornyakov

06:27 AM Feature #6187: eliminate explicit buffer usage in can-find

In my opinion, the converted code will read better with @canFind(...)@ than @new FindQuery(...).hasAny()@. Greg Shah

06:23 AM Feature #6187: eliminate explicit buffer usage in can-find

We really should just implement a set of static @canFind()@ methods which have the @<dmo>.Buf.class@ as a parameter a... Greg Shah

04:30 AM Feature #6187: eliminate explicit buffer usage in can-find

This is still an issue. It can be see with this simple test:... Constantin Asofiei

03/24/2022

03:22 PM Bug #3896: improve performance of the _lock metadata implementation

@LockTableUpdater@ is re-worked in the same way as @UserTableStatUpdater@ was re-worked some time ago. No separated t... Igor Skornyakov

03/18/2022

08:26 AM Feature #6187: eliminate explicit buffer usage in can-find

It seems in 2015, this may already have been resolved (from @record_scoping_prep.rules@):... Greg Shah

03/17/2022

12:38 PM Bug #3896: improve performance of the _lock metadata implementation

Does it make sense to move @LockTableUpdater.UpdateWorker.fileNums@ map to the @MetadataManager@ (as a value of the m... Igor Skornyakov

07:24 AM Feature #6187: eliminate explicit buffer usage in can-find

I don't recall why we require this extra buffer that is not actually a buffer in the 4GL. I do recall how painful th... Greg Shah

07:23 AM Feature #6187 (New): eliminate explicit buffer usage in can-find

Greg Shah

03/13/2022

04:47 AM Bug #3896 (WIP): improve performance of the _lock metadata implementation

Igor Skornyakov

03/04/2022

07:32 AM Bug #6145: "Bad" characters in the query.

It is possible that some characters in the string literals or string parameters of the query are not compatible with ... Igor Skornyakov

07:26 AM Bug #6145 (New): "Bad" characters in the query.

Igor Skornyakov