FWD » Core Development » Database

12/04/2025

05:35 PM Bug #10937 (Review): fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

I rebased and committed the final changes to FWD/10937a and H2/10937a_h2 branches. Now the testcases seem to work as ... Ovidiu Maxiniuc

08:30 AM Feature #4369 (WIP): implement stateless FWD server clustering

Greg Shah

07:52 AM Bug #9633 (Review): bulk delete for permanent tables

Stefanel Pezamosca

07:50 AM Bug #9633: bulk delete for permanent tables

I rebased 9633a to trunk revision 16302.
I have committed to *9633a revision 16305*:
* Reverted @ignore_scope... Stefanel Pezamosca

12/03/2025

11:02 AM Feature #9577: persist the artifacts from dynamic query parsing/conversion (including the JAST) a...

In addition, from today discussion, we would want to add a 'quality' property: save only converted queries which have... Ovidiu Maxiniuc

08:29 AM Bug #10962: Optimized multi-table compound components are not invalidated

FYI, there is @CQ.stateChanged@ that sets @useSnapshot@, @initReferenceRowSupport@ and @SnapshotParameterFilter@ to b... Alexandru Lungu

08:17 AM Bug #10962: Optimized multi-table compound components are not invalidated

I went to ~rev. 15550 but this is still reproducible. I still tend to think this is a regression and this worked at s... Alexandru Lungu

07:35 AM Bug #10962: Optimized multi-table compound components are not invalidated

+ Stefanel Alexandru Lungu

07:34 AM Bug #10962: Optimized multi-table compound components are not invalidated

It is not enough apparently. Even if I force the multi-table AQ to invalidate, it will get to the absolute first row.... Alexandru Lungu

07:17 AM Bug #10962: Optimized multi-table compound components are not invalidated

There is an issue with the invalidation of @AdaptiveQuery@ instances that are generated by the optimizer. It seems li... Alexandru Lungu

07:10 AM Bug #10962 (WIP): Optimized multi-table compound components are not invalidated

Alexandru Lungu

07:16 AM Bug #10956: Wrong primary index selection on FOR EACH table

Teodor Gorghe wrote:
> Created task branch *10956a*
>
> Committed revision *16299* on task branch *10956a*:
> - ... Constantin Asofiei

04:39 AM Bug #10956 (Review): Wrong primary index selection on FOR EACH table

Created task branch *10956a*
Committed revision *16299* on task branch *10956a*:
- Don't copy the @KW_PRIMARY@ pr... Teodor Gorghe

07:07 AM Feature #10959: Improve performance of dynamic queries

I think the most part of the queries which causes the cache to fill are more simple, like:... Ovidiu Maxiniuc

07:02 AM Feature #10959: Improve performance of dynamic queries

> How do you know where pred has been built so that we use this special structure instead of concatenating? That's wh... Alexandru Lungu

06:35 AM Feature #10959: Improve performance of dynamic queries

Some class has this method:... Constantin Asofiei

06:24 AM Feature #10959: Improve performance of dynamic queries

... Alexandru Lungu

06:19 AM Feature #10959: Improve performance of dynamic queries

Alexandru Lungu wrote:
> In my idea, the key would look like: @for each tt where tt.f1 = <token> and tt.f2 = <token>... Constantin Asofiei

06:12 AM Feature #10959: Improve performance of dynamic queries

> An idea is to have a minimal parser which has two purposes:
This is why I think we can rely on the converted co... Alexandru Lungu

05:57 AM Feature #10959: Improve performance of dynamic queries

Alex, whatever we do, the predicate string still needs to be parsed. An idea is to have a minimal parser which has t... Constantin Asofiei

05:34 AM Feature #10959: Improve performance of dynamic queries

Ovidiu, there are also some patterns quite often encountered, like:
* @hq:query-prepare(substitute("for each tt wh... Alexandru Lungu

12/02/2025

03:48 PM Feature #10959: Improve performance of dynamic queries

Recent work performed by FWD revealed a flaw in the implementation of 1st level cache of the dynamic query parser. Be... Ovidiu Maxiniuc

03:24 PM Feature #10959 (New): Improve performance of dynamic queries

Ovidiu Maxiniuc

07:47 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

I addressed the no-undo issue on @doCloseMultiplexScope@. It seems to eliminate the crash, leaving the table in a usa... Ovidiu Maxiniuc

07:23 AM Feature #4369: implement stateless FWD server clustering

Greg Shah wrote:
> Please still work to put as much of the Geode-specific logic into a clustering package. If possi... Andrei Plugaru

07:17 AM Feature #4369: implement stateless FWD server clustering

I understand that the logic in various places will need to be changed to be compatible with a synchronization approac... Greg Shah

06:08 AM Feature #4369: implement stateless FWD server clustering

The core data structures we need are mainly already implemented by Geode. They have @Region@, which is basically a di... Andrei Plugaru

06:59 AM Bug #10956 (WIP): Wrong primary index selection on FOR EACH table

I will assign this to me and I will work on this in parallel. Teodor Gorghe

06:50 AM Bug #10956: Wrong primary index selection on FOR EACH table

OK, so this is a conversion issue: if the model table has a primary index, and the definition also has one, then the ... Constantin Asofiei

06:38 AM Bug #10956: Wrong primary index selection on FOR EACH table

Result of @WRITE-XMLSCHEMA@:
* 4GL... Teodor Gorghe

06:29 AM Bug #10956: Wrong primary index selection on FOR EACH table

Teodor, the goal of WRITE-XML with schema included is to see which indexes are reported. Please post the .xml file h... Constantin Asofiei

06:23 AM Bug #10956: Wrong primary index selection on FOR EACH table

Confirmed:
* the order of WRITE-JSON/WRITE-XML is the same as for the @FOR EACH@ execution:
WRITE-JSON and WRITE-... Teodor Gorghe

06:12 AM Bug #10956: Wrong primary index selection on FOR EACH table

The customer uses @WRITE-JSON@, which the order is not the same, because of this issue.
I am testing with @WRITE-J... Teodor Gorghe

06:10 AM Bug #10956: Wrong primary index selection on FOR EACH table

Teodor, do a @WRITE-XML@ on tt2, with schema included, and see what indexes are reported there. Constantin Asofiei

05:25 AM Bug #10956: Wrong primary index selection on FOR EACH table

Suppose we have the following simple case:... Teodor Gorghe

05:20 AM Bug #10956 (Closed): Wrong primary index selection on FOR EACH table

Teodor Gorghe

12/01/2025

11:22 AM Feature #4369: implement stateless FWD server clustering

I'll have to think about the pattern a bit more before I discuss that.
I do want us to avoid putting the same Geod... Greg Shah

11/28/2025

10:10 AM Feature #4369: implement stateless FWD server clustering

I have been lately focused on implementing refactoring the @LockStatus/TraceLockStatus@ classes in order to be able t... Andrei Plugaru

09:29 AM Support #10664: Develop a multi session test suite in harness

Artur Școlnic wrote:
> Tomasz Domin wrote:
> > I fully agree, I was just wondering if I can help with tooling modif... Tomasz Domin

08:35 AM Support #10664: Develop a multi session test suite in harness

Tomasz Domin wrote:
> I fully agree, I was just wondering if I can help with tooling modification.
Since you brou... Artur Școlnic

08:27 AM Support #10664: Develop a multi session test suite in harness

Artur Școlnic wrote:
> We know there are failures, hence the skewed results, we simply need to work on the issues an... Tomasz Domin

08:26 AM Support #10664: Develop a multi session test suite in harness

We know there are failures, hence the skewed results, we simply need to work on the issues and bring the success rate... Artur Școlnic

08:18 AM Support #10664: Develop a multi session test suite in harness

Artur Școlnic wrote:
> I finished writing the locking tests, the results are:
> OE: 1610/1610
> FWD: 411/1610
>... Tomasz Domin

05:19 AM Support #10664 (Review): Develop a multi session test suite in harness

I finished writing the locking tests, the results are:
OE: 1610/1610
FWD: 411/1610
There are two main reasons wh... Artur Școlnic

03:47 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Committed revision *16298* on task branch *10455b*:
- Set the batch mode validation to false when it is in bulk copy. Teodor Gorghe

01:49 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Committed revision *16297* on task branch *10455b*:
- Changed reportBatchValidationException to throw the error. Teodor Gorghe

01:43 AM Bug #10455: Persistent table CAN-FIND FIRST issue

I am thinking on something else:
* when the assigned buffer is the destination buffer, or something else, a variable... Teodor Gorghe

01:05 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Constantin Asofiei wrote:
> Also, does the original test scenario still work?
Yes, they still work, I have tested... Teodor Gorghe

11/27/2025

12:09 PM Bug #10455: Persistent table CAN-FIND FIRST issue

For show vs throwError I think there are two cases:
* if validation fails and LVALUE is the buffer (or field?) which... Constantin Asofiei

10:50 AM Bug #10455 (Review): Persistent table CAN-FIND FIRST issue

Committed revision *16296* on task branch *10455b*:
- Fixed BUFFER-COPY BATCH ASSIGN when a function call is the fir... Teodor Gorghe

09:50 AM Bug #10455: Persistent table CAN-FIND FIRST issue

I have also tried this:... Teodor Gorghe

09:38 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Hm... you are right, this works. But, how can we differentiate that we are in an assignment for a target buffer fiel... Constantin Asofiei

09:33 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Please try this:... Teodor Gorghe

09:29 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Note that @ErrorManager.recordOrThrowError(newErrorCodes, newErrorMessages, false);@ is called in @RecordBuffer.repor... Constantin Asofiei

09:26 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Teodor, please check with 12.x this test:... Constantin Asofiei

09:01 AM Bug #10455: Persistent table CAN-FIND FIRST issue

Created task branch *10455b*. Teodor Gorghe

09:00 AM Bug #10455 (WIP): Persistent table CAN-FIND FIRST issue

There is a regression regarding this task.
When the @BUFFER-COPY ASSIGN@ on a buffer on the same table with unique f... Teodor Gorghe

08:08 AM Bug #6834 (Closed): RAQ navigation on records having NULL values on same index components

Dănuț Filimon

08:06 AM Bug #6834: RAQ navigation on records having NULL values on same index components

Cool. Then this can be closed. Ovidiu Maxiniuc

06:25 AM Bug #6834: RAQ navigation on records having NULL values on same index components

This has been fixed in a revision before trunk/16000.
The FWD output of the #6834-3 test is:... Dănuț Filimon

05:20 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

> We can TRUNCATE only if there are separate SQL tables, and we get rid of the MULTIPLEX. Otherwise, we need DELETE .... Alexandru Lungu

05:05 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

Alexandru Lungu wrote:
> Mind that this does TRUNCATE AFAIK.
We can TRUNCATE only if there are separate SQL tables,... Constantin Asofiei

04:51 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

> Otherwise, if it proves too difficult to solve/investigate the destructor conundrums, lets work on enhancing the UP... Alexandru Lungu

04:41 AM Feature #9711: collation configuration cleanup

Stefanel Pezamosca wrote:
> If I understand correctly, the goal here is to to eliminate the need for manual configur... Alexandru Lungu

03:26 AM Feature #9711: collation configuration cleanup

If I understand correctly, the goal here is to to eliminate the need for manual configuration and hard-coded settings... Stefanel Pezamosca

11/26/2025

02:48 PM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

Ovidiu, lets solve the @nonEmptyTableFlags@ issue in this task:
* for NO-UNDO tables, this bitmap will work as is - ... Constantin Asofiei

08:44 AM Bug #10937 (WIP): fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

Ovidiu Maxiniuc

08:23 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

Or maybe the destructor is deferred till the end of the transaction and we don't see that. Maybe a message after dele... Alexandru Lungu

08:22 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

> In the temp-table destroy case, the destructor changes are not rolled back.
So you can create things in the pers... Alexandru Lungu

08:01 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

Alexandru Lungu wrote:
> > So, does an UNDO revert any created records by such destructors?
>
> Huh, interesting... Constantin Asofiei

05:07 AM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

> I don't see another way than add a NO-UNDO option for the DELETE statement in H2, so that we can do DELETEs which a... Alexandru Lungu

11/25/2025

01:35 PM Bug #10937: fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

This is related also to the #10833-80 scenario, where it was proved that @nonEmptyTableFlags@ is broken in case of ro... Constantin Asofiei

01:16 PM Bug #10937 (Test): fwd-h2 'killedMultiplexesCleanup' must not abend if the batch is not empty

Constantin Asofiei

06:23 AM Bug #10901 (Review): BEGINS trims the rvalue if the lvalue field is first in an index

Created 10901a from trunk rev 16292. In rev 16293: If in a BEGINS LVALUE is a field part of the index used for sort,... Constantin Asofiei

11/24/2025

09:58 AM Feature #6348: implement support for MariaDB

It should be a different task, please make one in Deployment. Greg Shah

09:49 AM Feature #6348: implement support for MariaDB

Is it fair to continue use of this task for finishing touches on MariaDB? Some things I am considering:
* We have im... Roger Borrello

06:28 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

> Is not about a query from FWD, just a direct SELECT I noticed. Yes, RTRIM=TRUE is in the JDBC URL.
Right. The co... Alexandru Lungu

06:24 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

Is not about a query from FWD, just a direct SELECT I noticed. Yes, @RTRIM=TRUE@ is in the JDBC URL. Constantin Asofiei

04:31 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

PS: I see @dialect.addRtrimmedExpression("?", sb);@. This can be wrapped with @dialect.isAutoRtrim@. Alexandru Lungu

04:30 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

Is @RTRIM=TRUE;@ set? H2 also knows to compare the values of strings with rtrim automatically. Alexandru Lungu

04:15 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

Alex, as a side note: this query does not return records in H2: @select * from meta_user where userid = rtrim('HOTEL'... Constantin Asofiei

03:35 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

I am OK with the changes in #10920-3. This piece of code is a backdoor into the persistence and wasn't handled proper... Alexandru Lungu

03:23 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

This is because of the H2's dialect automatic case-insensitive assumption in @P2JH2Dialect@; I've used a JDBC URL li... Constantin Asofiei

11/23/2025

05:42 AM Bug #10920: String concatenation doesn't work in WHERE clause for permanent databases

I used H2 permanent database. No records are found in FWD.... Stanislav Lomany

05:40 AM Bug #10920 (New): String concatenation doesn't work in WHERE clause for permanent databases

Stanislav Lomany

11/21/2025

09:48 AM Feature #4369: implement stateless FWD server clustering

Greg Shah wrote:
> The tables we could expect to be used frequently are not the same tables that are editable. @_fi... Eric Faulhaber

08:09 AM Feature #4369: implement stateless FWD server clustering

The tables we could expect to be used frequently are not the same tables that are editable. @_file@, @_field@, @_ind... Greg Shah

06:09 AM Feature #4369: implement stateless FWD server clustering

I am just writing here a discussion that happened in the daily yesterday.
Back when we tested performance for a la... Alexandru Lungu

11/20/2025

01:55 PM Feature #4369: implement stateless FWD server clustering

> In the end, it all comes down to the number of queries executed for the metadata tables for the current and future ... Greg Shah

12:52 PM Feature #4369: implement stateless FWD server clustering

Today I have made some experiments for understanding the time taken for inserting a large number of records in a meta... Andrei Plugaru

11/19/2025

02:07 PM Feature #4369: implement stateless FWD server clustering

> Over the solution presented by you, this has the advantage that the tables don't need to be recreated before each q... Greg Shah

01:14 PM Feature #4369: implement stateless FWD server clustering

Greg Shah wrote:
> For the tables that need to represent state that can change at runtime, I am suggesting that the ... Andrei Plugaru

11:02 AM Feature #4369: implement stateless FWD server clustering

> Greg, just to be clear, do you suggest to give up on the H2 metadata tables all together?
No. There is value i... Greg Shah

06:52 AM Feature #4369: implement stateless FWD server clustering

Greg Shah wrote:
> I repeat my points of #4369-46 and #4369-54. The existing approach to metadata is overcomplicate... Andrei Plugaru

11/18/2025

05:36 PM Feature #4369: implement stateless FWD server clustering

I repeat my points of #4369-46 and #4369-54. The existing approach to metadata is overcomplicated and causes unneces... Greg Shah

04:52 PM Feature #4369: implement stateless FWD server clustering

Andrei Plugaru wrote:
> [...]
> IMHO, the versioning approach from @UserTableStatUpdater@ and @LockTableUpdater@ is... Eric Faulhaber

10:32 AM Feature #4369: implement stateless FWD server clustering

In the last note, I forgot to mentioned the new directory configurations introduced in my latest commits. Right under... Andrei Plugaru

07:42 AM Feature #4369: implement stateless FWD server clustering

I have committed @4369a/16232@. I have completed all the java docs, history entries all the file headers on the new f... Andrei Plugaru

11:34 AM Bug #10901: BEGINS trims the rvalue if the lvalue field is first in an index

Actually things are a little more complicated: what matters is that the field is part of the 4GL index associated wit... Constantin Asofiei

09:52 AM Bug #10901: BEGINS trims the rvalue if the lvalue field is first in an index

This test shows that if a field is part of a leading component, then it will trim the rvalue for @BEGINS@:... Constantin Asofiei

09:45 AM Bug #10901 (Closed): BEGINS trims the rvalue if the lvalue field is first in an index

Constantin Asofiei

11/17/2025

04:30 AM Bug #10890: merge-row-changes does not change the before/after image row-state correctly

Another issue is that after @merge-row-changes@, in OE there are scenarios in which the *before image is not deleted*... Artur Școlnic

04:27 AM Bug #10890: merge-row-changes does not change the before/after image row-state correctly

... Artur Școlnic

04:19 AM Bug #10890 (New): merge-row-changes does not change the before/after image row-state correctly

Artur Școlnic

11/15/2025

08:28 AM Feature #4369: implement stateless FWD server clustering

I have mainly finished all the changes necessary in @InMemoryLockManager@ regarding the @.wait()/.notify()@ construct... Andrei Plugaru

11/14/2025

05:56 AM Feature #9711: collation configuration cleanup

Alex, when talking _meta tables, the nuances you mentioned is a bit more complex.
There is some 'persistent' infor... Ovidiu Maxiniuc

05:33 AM Feature #9711: collation configuration cleanup

> I think we need a way to allow the H2's default collation and ignore completely the embedded-collation for the _tem... Alexandru Lungu

04:58 AM Feature #9711: collation configuration cleanup

I think we need a way to allow the H2's default collation and ignore completely the @embedded-collation@ for the _tem... Constantin Asofiei

11/13/2025

07:44 AM Bug #10385 (WIP): Incorrect reset of the cursor cache in Reposition

I run a large unit test suite and discovered a NPE in @RepositionCache$Node.getRow@, on @rows.floor(position.intValue... Alexandru Lungu

05:52 AM Feature #4369: implement stateless FWD server clustering

> If the requests are stateless, then they should be able to be distributed across the clustered servers without any ... Alexandru Lungu

11/12/2025

10:56 AM Feature #4369: implement stateless FWD server clustering

> if there are 1k requests at a time and 10 servers, it is completely valid to balance 100 across these 10 servers?
... Greg Shah

10:54 AM Feature #4369: implement stateless FWD server clustering

I don't think we need the @RemoteLockManager@ if we have proper clustering. Greg Shah

09:15 AM Feature #4369: implement stateless FWD server clustering

> Within single-server mode and cluster mode, the server should still handle multiple requests in parallel? I mean, i... Andrei Plugaru

07:59 AM Feature #4369: implement stateless FWD server clustering

Review of 4369a:
h4. Model
* I spot some changes in the synchronization in general and I want to better underst... Alexandru Lungu

11/10/2025

08:41 AM Feature #4369: implement stateless FWD server clustering

Queuing it up! Alexandru Lungu

08:33 AM Feature #4369 (Review): implement stateless FWD server clustering

Eric/Alex: Please do this first review. Greg Shah

07:35 AM Bug #10838: Unique index violation should be cured even at flush time

Make a @tbl@ persistent table with unique index of (f1, f2).... Alexandru Lungu

07:32 AM Bug #10838 (New): Unique index violation should be cured even at flush time

Alexandru Lungu

11/07/2025

08:46 AM Bug #9950 (Review): Upgrade the metadata schema to a newer version

Constantin Asofiei

08:34 AM Feature #9305: Replace begins with interval based conditional

Please do the tests requested in #9305-25 in 4GL and PG with LIKE / < and > / with BETWEEN. We need to take a decisio... Alexandru Lungu

07:42 AM Feature #9305: Replace begins with interval based conditional

The range check follows the field collation or is byte-by-byte?
If it is after collation, there are some cases like ... Teodor Gorghe

07:11 AM Feature #9305: Replace begins with interval based conditional

Teodor Gorghe wrote:
> You told yesterday and in #9305-10 that the default database configuration, with @en_us.UTF... Eduard Soltan

07:01 AM Bug #7999: FWD does not honor FIELDS/EXCEPT at dynamic queries

When the changes from this task are regression tested, please also include the following scenario mentioned in #10763... Lorian Sandu

11/06/2025

04:55 PM Bug #9950: Upgrade the metadata schema to a newer version

I added backward compatibility with projects that have not upgraded from OE 11.6. By default 11.6 is the default meta... Ovidiu Maxiniuc

10:45 AM Feature #4369 (WIP): implement stateless FWD server clustering

*Update on the current status.*
Up until now, I have synchronized the *tracker for records in unique indices* and th... Andrei Plugaru

05:36 AM Feature #9305: Replace begins with interval based conditional

Alexandru,
You told yesterday and in #9305-10 that the default database configuration, with @en_us.UTF8@ collation, ... Teodor Gorghe

04:04 AM Bug #10822: Lock is not acquired if the update field value is the same as the current one

If a no-lock is set on a record it cannot be modified, 4GL does not allow any assignment and raises ... Artur Școlnic

03:58 AM Bug #10822 (New): Lock is not acquired if the update field value is the same as the current one

Artur Școlnic

03:56 AM Support #10664: Develop a multi session test suite in harness

I agree, I will open related tasks for every issue found during testing. Artur Școlnic

03:53 AM Support #10664: Develop a multi session test suite in harness

> Alex/Eric can determine if we do this now or if we do this in a separate task.
IMHO, related tasks should be ope... Alexandru Lungu

02:34 AM Support #10664: Develop a multi session test suite in harness

Artur Școlnic wrote:
> The tests that are not running properly have this pattern:
> [...]
>
> The test relies on... Greg Shah

02:12 AM Support #10664: Develop a multi session test suite in harness

Razvan pointed out an issue with lock scopes in #10796, @locking@ test suite already includes lock release/downgrade ... Artur Școlnic