Add OL facet tables #2152

wslulciuc · 2022-09-30T16:30:49Z

Problem

To access OL facets, we have to run direct queries against OL events table, which has significant performance concerns.

Closes: #2076

Solution

Add tables for dataset, job, and run facets.

Checklist

You've signed-off your work
Your changes are accompanied by tests (if relevant)
Your change contains a small diff and is self-contained
You've updated any relevant documentation (if relevant)
You've updated the CHANGELOG.md with details about your change under the "Unreleased" section (if relevant, depending on the change, this may not be necessary)
You've versioned your .sql database schema migration according to Flyway's naming convention (if relevant)
You've included a header in any source code files (if relevant)

pawel-big-lebowski

This is going to be huge improvement in terms of performance.
It will also help developing some features like getting raw events by namespace which @mobuchowski coped with some time ago (#2070)

My concerns are:

If my understanding is correct, current behavior leads users to losing dataset history (facets written the old way won't be read the new way). If this acceptable, at least we should put some doc notes. Otherwise, providing some migration (even the optional migration) would be great.
We're storing the same facets twice now: in lineage_events and new facets' tables. Is this an expected behavior?
Wouldn't we like to have a test that verifies after posting lineage event that introduced tables are filled? I couldn't spot anything similar in the PR.
Flyway DB migration version files numeration need to be adjusted (v46 seem to be outdated)

pawel-big-lebowski · 2022-12-05T13:44:29Z

api/src/main/java/marquez/db/DatasetDao.java

          + "        WHERE run_uuid = dv.run_uuid\n"
-          + "    ) e ON e.run_uuid = dv.run_uuid\n"
+          + "    ) df ON df.run_uuid = dv.run_uuid\n"


this seems to be missing dataset_uuid = dv.dataset_uuid

codecov · 2022-12-05T14:07:28Z

Codecov Report

Merging #2152 (ad0fa19) into main (ae6250a) will increase coverage by 0.41%.
The diff coverage is 81.64%.

@@             Coverage Diff              @@
##               main    #2152      +/-   ##
============================================
+ Coverage     76.72%   77.14%   +0.41%     
- Complexity     1177     1237      +60     
============================================
  Files           222      227       +5     
  Lines          5354     5727     +373     
  Branches        429      465      +36     
============================================
+ Hits           4108     4418     +310     
- Misses          768      801      +33     
- Partials        478      508      +30

Impacted Files	Coverage Δ
api/src/main/java/marquez/MarquezApp.java	`63.75% <0.00%> (-2.49%)`	⬇️
...src/main/java/marquez/cli/V55MigrationCommand.java	`0.00% <0.00%> (ø)`
api/src/main/java/marquez/db/DatasetDao.java	`98.64% <ø> (ø)`
api/src/main/java/marquez/db/JobDao.java	`100.00% <ø> (ø)`
api/src/main/java/marquez/db/JobVersionDao.java	`91.04% <ø> (ø)`
api/src/main/java/marquez/db/RunDao.java	`92.50% <ø> (ø)`
...ain/java/marquez/db/mappers/DatasetDataMapper.java	`85.18% <ø> (ø)`
...rc/main/java/marquez/db/mappers/JobDataMapper.java	`87.50% <ø> (ø)`
.../java/marquez/db/models/ColumnLineageNodeData.java	`100.00% <ø> (ø)`
.../src/main/java/marquez/service/LineageService.java	`80.30% <ø> (ø)`
... and 11 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

mobuchowski · 2022-12-21T13:06:01Z

metadata.json - do we need this file? It shows one line changed, but also 1 addition, 0 deletions not shown because the diff is too large. Please use a local Git client to view these changes.

mobuchowski · 2022-12-21T12:22:37Z

api/src/main/java/marquez/db/migrations/V55_5__BackfillFacets.java

+  private static final String GET_CURRENT_LOCK_SQL =
+      """
+    SELECT * FROM v55_facet_migration_lock ORDER BY created_at ASC, run_uuid ASC LIMIT 1
+  """;


I think oneliners need not to use text block syntax, especially if not well-formatted 😉

mobuchowski · 2022-12-21T12:53:47Z

api/src/main/java/marquez/db/RunDao.java

-          + "    SELECT le.run_uuid, JSON_AGG(event->'run'->'facets') AS facets\n"
-          + "    FROM lineage_events le\n"
-          + "    GROUP BY le.run_uuid\n"
+          + "    SELECT rf.run_uuid, JSON_AGG(rf.facet ORDER BY rf.lineage_event_time ASC) AS facets\n"


It's risky to use this data in the same version that we're migrating in, since we have manual migration process for some.

This release should IMO still use lineage_events table, and given reasonable time to migrate to newer version, we should use new tables to optimize queries.

Unless we have a well documented a way to run migration without impacting "running deployment", so that we don't have to rollback to previous Marquez version if something fails.

Thanks @mobuchowski for the pointing this out (which, I agree with).

@pawel-big-lebowski I know this PR / optimization has been dragging (mostly my fault, so happy to take the blame), but splitting the changes into separate PRs / releases would reduce any challenges with rollbacks, but also provide users the option to decide when to apply the migration / upgrade. So, that said, what I propose is:

Open a PR to just introduce the dataset, job, and run facets tables (a write only mode), meaning the facet tables won't be referenced in anyway in our queries, but will be written to so that features can take advantage of the optimization; lineage events will continue to be written to the lineage_events table as well. This can be released in 0.30.0.

As a follow up PR, we can update all references to lineage_events in our queries to use the individual facet tables. Within this PR, the migration script (and documentation) to backfill facets from the lineage_events table will also be introduced (and be required to run for users with larger lineage_events tables). This can be released in 0.31.0 or later.

PR #2152 (current) can be step 2 and marked ias blocked, and I can cherry pick any changes into a separate PR for step 1 (also open to other options). Anyways, thanks for the great work here @pawel-big-lebowski. Thoughts?

It may be difficult to split the change into separate PRs. The initial PR was missing tests. I added them within my commits and it detected some issues with previous commits like:

runFacetExists method in api/src/main/java/marquez/db/RunFacetsDao.java

fromName method in api/src/main/java/marquez/db/DatasetFacetsDao.java
(see my latest commit)

So merging some of the commits shouldn't be a solution. Additionally, one cannot take benefit from facet tables unless they are filled with data. Otherwise, users will experience incomplete results.

This PR does not clean or remove data from lineage_events table so we always have the ability to prepare a fix that restores previous mechanics if we want to.

@wslulciuc Although splitting this PR into multiple PRs could improve readability, I don't see any advantage of splitting this into separate releases (0.30.0 and 0.31.0).

The combination of a manual migration process and the fact that the code stops reading from the lineage_events table introduces a real problem for users. When a user deploys, the tables will be created and all new records will be written into the new facet tables. However, all old data still lives only in the lineage_events table and won't be accessible until after the user executes the migration command. This means that the service will effectively be broken until users choose to execute this migration.

I think splitting the PR so that writes and reads are committed and can be deployed separately makes sense. This gives users the opportunity to push multiple deployments in a way that guarantees no downtime. It also makes this PR muuuch easier to review - readability is not just good for its own sake; it makes it easier to catch issues before they go to production. The migration script itself warrants its own review, I think.

api/src/main/java/marquez/db/RunFacetsDao.java

api/src/test/java/marquez/db/migrations/V55_5__BackfillFacetsTest.java

api/src/main/java/marquez/db/migrations/V55_5__BackfillFacets.java

Signed-off-by: wslulciuc <[email protected]>

Signed-off-by: Pawel Leszczynski <[email protected]>

pawel-big-lebowski · 2023-01-02T09:49:21Z

Database migration in a nutshell:

Users with up to 100K lineage_events should not bother about this migration.
Users with more than 100K lineage_events will need to run additional migration command.
DB upgrade is millisecond long. This is because migration procedure creates views with job, run and dataset facets. When target facet tables are loaded, the views are pointing to them. Until that moment views are defined over lineage_events table.
After the change facets are written into two locations: lineage_events table and job/run/dataset facets tables.
Perf tests were run to verify data migration performance. See results.

wslulciuc · 2023-01-04T09:43:01Z

metadata.json - do we need this file? It shows one line changed, but also 1 addition, 0 deletions not shown because the diff is too large. Please use a local Git client to view these changes.

I updated metadata.json under docker/ to include missing facets, but looks like the other meta file was checked it by mistake and can be removed.

collado-mike · 2023-01-05T00:23:30Z

api/src/main/java/marquez/db/DatasetFacetsDao.java

+    Optional.ofNullable(datasetFacets.getDocumentation())
+        .ifPresent(
+            documentation ->
+                insertDatasetFacet(
+                    UUID.randomUUID(),
+                    now,
+                    datasetUuid,
+                    runUuid,
+                    lineageEventTime,
+                    lineageEventType,
+                    Facet.DOCUMENTATION.getType(),
+                    Facet.DOCUMENTATION.getName(),
+                    toPgObject(Facet.DOCUMENTATION.asJson(documentation))));


You can replace the copy/paste of this block with something like

writeDatasetFacet(Facet.DOCUMENTATION, datasetFacets.getDocumentation());

where writeDatasetFacet is defined as

void writeDatasetFacet(Facet facet, Object nullableData) { Optional.ofNullable(nullableData) .ifPresent( data -> insertDatasetFacet( UUID.randomUUID(), now, datasetUuid, runUuid, lineageEventTime, lineageEventType, facet.getType(), facet.getName(), toPgObject(facet.asJson(data)))); }

I created separate PR #2350 which is a part of this one responsible for creating new tables, writing new events to new tables, while not reading data from new tables.

I got rid of the repeated code block in #2350

collado-mike · 2023-01-05T00:25:31Z

api/src/main/java/marquez/db/JobFacetsDao.java

+    Optional.ofNullable(jobFacet.getDocumentation())
+        .ifPresent(
+            documentation ->
+                insertJobFacet(
+                    UUID.randomUUID(),
+                    now,
+                    jobUuid,
+                    runUuid,
+                    lineageEventTime,
+                    lineageEventType,
+                    DatasetFacetsDao.Facet.DOCUMENTATION.getName(),
+                    toPgObject(DatasetFacetsDao.Facet.DOCUMENTATION.asJson(documentation))));
+


An interface for the various facet enums will help you avoid all this duplication

I agree, it's a bit ugly. Do you mean defining an interface for each enum to implement?

I created separate PR #2350 which is a part of this one responsible for creating new tables, writing new events to new tables, while not reading data from new tables.

I addressed this feedback remark in #2350.

collado-mike · 2023-01-05T01:21:47Z

api/src/main/java/marquez/db/RunDao.java

-          + "    SELECT le.run_uuid, JSON_AGG(event->'run'->'facets') AS facets\n"
-          + "    FROM lineage_events le\n"
-          + "    GROUP BY le.run_uuid\n"
+          + "    SELECT rf.run_uuid, JSON_AGG(rf.facet ORDER BY rf.lineage_event_time ASC) AS facets\n"


The combination of a manual migration process and the fact that the code stops reading from the lineage_events table introduces a real problem for users. When a user deploys, the tables will be created and all new records will be written into the new facet tables. However, all old data still lives only in the lineage_events table and won't be accessible until after the user executes the migration command. This means that the service will effectively be broken until users choose to execute this migration.

I think splitting the PR so that writes and reads are committed and can be deployed separately makes sense. This gives users the opportunity to push multiple deployments in a way that guarantees no downtime. It also makes this PR muuuch easier to review - readability is not just good for its own sake; it makes it easier to catch issues before they go to production. The migration script itself warrants its own review, I think.

pawel-big-lebowski · 2023-01-05T07:46:58Z

Replying to @collado-mike

The combination of a manual migration process and the fact that the code stops reading from the lineage_events table introduces a real problem for users. When a user deploys, the tables will be created and all new records will be written into the new facet tables. However, all old data still lives only in the lineage_events table and won't be accessible until after the user executes the migration command.

I think there may be some kind of misunderstanding here. For the users with more than 100K lineage_events:

run_facets, job_facets and dataset_facets tables are created during the flyway migration,
Views run_facets_view, job_facets_view and dataset_facets_view are created during flyway migration and their definition is created on the top of lineage_events tables. Application code makes use of those views.
Filling run_facets, job_facets and dataset_facets with data has to be done as manual asynchronous step but the application is still up and running before the command is called and executed.
Once manual migration is finished, views run_facets_view, job_facets_view and dataset_facets_view are recreated and their definition is based on run_facets, job_facets and dataset_facets tables.

This means that the service will effectively be broken until users choose to execute this migration.

No, it's not and that's why run_facets_view, job_facets_view and dataset_facets_view are created. They provide no-downtime experience at the time of manual migration procedure.

I think splitting the PR so that writes and reads are committed and can be deployed separately makes sense. This gives users the opportunity to push multiple deployments in a way that guarantees no downtime. It also makes this PR muuuch easier to review - readability is not just good for its own sake; it makes it easier to catch issues before they go to production. The migration script itself warrants its own review, I think.

I agree that splitting PR into multiple PRs makes sense:

PR I - create run_facets, job_facets and dataset_facets tables, write to them for new OL events. don't read data from new tables.
PR II - create run_facets_view, job_facets_view and dataset_facets_view on the top of lineage_events table, switch application logic to making use of the views.
PR III - migration command to fill run_facets, job_facets and dataset_facets tables. Replace the views from PR II to point to run_facets, job_facets and dataset_facets tables once migration is over.

Releasing PR I, PR II and PR III in one go does not imply any downtime.

wslulciuc · 2023-01-05T08:41:44Z

CHANGELOG.md

+* Split `lineage_events` table to `dataset_facets`, `run_facets`, and `job_facets` tables. [`2152`](https://github.com/MarquezProject/marquez/pull/2152) 
+[@wslulciuc](https://github.com/wslulciuc,), [@pawel-big-lebowski]( https://github.com/pawel-big-lebowski)  
+  * Performance improvement with migration procedure that requires manual steps if database has more than 100K lineage events.
+  * Please read [here](https://github.com/MarquezProject/marquez/blob/main/api/src/main/resources/marquez/db/migration/V55__readme.md) to get more database migration details.


Minor: I would reword as follows:

Note: We highly encourage users to review our migration plan.

wslulciuc · 2023-01-05T08:58:26Z

api/src/main/java/marquez/MarquezApp.java

@@ -149,6 +150,12 @@ public void registerResources(
    }
  }

+  @Override
+  protected void addDefaultCommands(Bootstrap<MarquezConfig> bootstrap) {


To avoid overriding addDefaultCommands(), we can just register the command within MarquezApp.initialize():

@Override public void initialize(@NonNull Bootstrap<MarquezConfig> bootstrap) { bootstrap.addCommand(new V55MigrationCommand()); // continue initialization ... }

The existing commands are kind of no-arg commands. Migration commands needs to be able to access database connection. That's why I included it here so that it can get Application as a constructor param. Within initialize commands are added although application is not created yet.

wslulciuc · 2023-01-05T09:15:45Z

api/src/main/java/marquez/cli/V55MigrationCommand.java

+ * <p>Please refer to @link marquez/db/migration/V55__readme.md for more details.
+ */
+@Slf4j
+public class V55MigrationCommand<MarquezConfig> extends EnvironmentCommand<marquez.MarquezConfig> {


Is V55MigrationCommand too specific (but also vague)? We'll certainly have more migrations in the future and one-off commands for each will be hard to manage and not the best user experience. I suggest we name the class DbMigrationsCommand and define v55_migrate as a subcommand. This tutorial is a pretty good reference on how to define subcommands.

wslulciuc · 2023-01-05T09:22:02Z

api/src/main/java/marquez/db/JobFacetsDao.java

+    Optional.ofNullable(jobFacet.getDocumentation())
+        .ifPresent(
+            documentation ->
+                insertJobFacet(
+                    UUID.randomUUID(),
+                    now,
+                    jobUuid,
+                    runUuid,
+                    lineageEventTime,
+                    lineageEventType,
+                    DatasetFacetsDao.Facet.DOCUMENTATION.getName(),
+                    toPgObject(DatasetFacetsDao.Facet.DOCUMENTATION.asJson(documentation))));
+


I agree, it's a bit ugly. Do you mean defining an interface for each enum to implement?

wslulciuc · 2023-01-05T09:28:35Z

api/src/main/java/marquez/db/migrations/V55_5__BackfillFacets.java

+
+  @Setter private Integer chunkSize = null;
+
+  @Setter private boolean triggeredByCommand = false;


Minor: I'd use the flag name manual.

wslulciuc · 2023-01-05T09:29:27Z

api/src/main/java/marquez/db/migrations/V55_5__BackfillFacets.java

+    execute(CREATE_DATASET_FACETS_VIEW);
+    execute(CREATE_JOB_FACETS_VIEW);
+
+    if (!triggeredByCommand && countLineageEvents() >= BASIC_MIGRATION_LIMIT) {


It's great that we inform the user on how to handle very large tables 💯

wslulciuc · 2023-01-05T09:32:25Z

api/src/main/java/marquez/db/migrations/V55_5__BackfillFacets.java

+     * Workaround to register uuid_generate_v4 function to generate uuids. gen_random_uuid() is
+     * available since Postgres 13
+     */
+    jdbi.withHandle(h -> h.createCall("CREATE EXTENSION IF NOT EXISTS \"uuid-ossp\"").invoke());


It might be possible that the migration might not be applied by a superuser? Can we the uuid using java.util.UUID, then inject the value in the inserts?

That's a hard one. So, we don't want Java to touch the data. It's good for performance when everything is done in SQLs and no data transfers between backend DB and Java application are required.

Adding this extension is a workaround for PostgreSQL 12, that we still support. If we stopped supporting PostgreSQL 12, we could use built-in uuid function (https://www.postgresql.org/docs/current/functions-uuid.html) available since version 13.

Otherwise we can add extra condition here to include extra extension only for PostgreSQL 12. Anyway, I think PostgreSQL 12 users will need a superuser.

Ok, that's reasonable. Mind making a note in the migration doc on the superuser requirement? Also, yeah, we should upgrade our version requirements for PostgreSQL. I wanted to first have in place schema validation in CI. PR #2326 gets us one step closer to upgrading.

wslulciuc · 2023-01-05T09:36:43Z

api/src/main/resources/marquez/db/migration/V55__readme.md

+| 10K events       | 50K rows     | 50K rows     | 150K rows        | 10sec      |
+| 100K events      | 500K rows    | 500K rows    | 150K rows        | 106sec     |
+| 500K events      | 2.5M rows    | 2.5M rows    | 7.5M rows        | 612sec     |
+| 1M events        | 5M rows      | 5M rows      | 15M rows         | 1473sec    |


Thanks for this detailed and well throughout migration plan @pawel-big-lebowski 💯

wslulciuc · 2023-01-05T10:01:55Z

The combination of a manual migration process and the fact that the code stops reading from the lineage_events table introduces a real problem for users.

In step 1 that I outlined above, Marquez would continue to write to the lineage_events table (and simultaneously to the new facet tables) therefore addressing this concern, which is a significant operational advantage. At Astronomer, we'd want this type of control to limit any downtime during the database migration. But, after doing another pass on the PR (and additional context from @pawel-big-lebowski comment above), views are used to avoid such downtime and effectively provide the safe guards for users when (manually, or automatically) applying the migration.

@pawel-big-lebowski has made some great additions and has helped push this PR to it's current state (and close to being merged!). Based on your feedback @collado-mike and @pawel-big-lebowski's input, there are enough safe guards to ensure the user has control of how the migration is applied (as outlined in V55__readme.md) while also ensuring Marquez can continue to handle reads / writes. To not completely stop progress, I feel addressing the comments in the PR and cleaning up a few sections of the code would avoid splitting this PR. Yeah, the PR is a bit on the larger size, but manageable and all additions are relevant to the feature. We can still split up the PR if that's more helpful for reviewing, but I don't think it's required. My comment earlier about separate releases of Marquez for this feature can be ignored.

pawel-big-lebowski · 2023-01-17T09:39:49Z

The content of this PR has been split to:

Review comments provided in this PR have been applied there.

boring-cyborg bot added the api API layer changes label Sep 30, 2022

mobuchowski requested a review from pawel-big-lebowski October 3, 2022 14:43

pawel-big-lebowski reviewed Oct 5, 2022

View reviewed changes

wslulciuc mentioned this pull request Nov 22, 2022

Only first job context is taken into consideration. #2230

Closed

pawel-big-lebowski reviewed Dec 5, 2022

View reviewed changes

pawel-big-lebowski force-pushed the feature/ol-facet-tables branch from bea5995 to e789723 Compare December 5, 2022 14:01

boring-cyborg bot added the docker label Dec 5, 2022

pawel-big-lebowski self-assigned this Dec 12, 2022

pawel-big-lebowski force-pushed the feature/ol-facet-tables branch 3 times, most recently from 2e850f8 to f1b5ffc Compare December 13, 2022 09:41

pawel-big-lebowski marked this pull request as draft December 13, 2022 09:48

pawel-big-lebowski force-pushed the feature/ol-facet-tables branch 8 times, most recently from 70471fb to 660f093 Compare December 20, 2022 09:08

boring-cyborg bot added the docs label Dec 20, 2022

pawel-big-lebowski force-pushed the feature/ol-facet-tables branch 3 times, most recently from 1b0ccd9 to 3830c32 Compare December 21, 2022 08:53

mobuchowski requested changes Dec 21, 2022

View reviewed changes

pawel-big-lebowski force-pushed the feature/ol-facet-tables branch 3 times, most recently from c2c16f5 to 1748433 Compare December 22, 2022 12:22

wslulciuc and others added 13 commits December 29, 2022 10:37

Add missing documentation and datasource to datasets in metadata.json

a52e9a1

Signed-off-by: wslulciuc <[email protected]>

continued: Add JobFacetsDao and DatasetFacetsDao

2c017a5

Signed-off-by: wslulciuc <[email protected]>

WIP

1ba2603

Signed-off-by: wslulciuc <[email protected]>

Add nominalTime to metadata.json and RunFacetsDao.insertRunFacetsFor()

879da87

Signed-off-by: wslulciuc <[email protected]>

Use run_facets table in RunDao

e48849a

Signed-off-by: wslulciuc <[email protected]>

Use run_facets in LineageDao.getCurrentRuns()

66eda9e

Signed-off-by: wslulciuc <[email protected]>

Define DatasetFacetsDao.Type

22141f4

Signed-off-by: wslulciuc <[email protected]>

Update job queries to use job_facets table

ec66136

Signed-off-by: wslulciuc <[email protected]>

Fix Facet.fromName() for job and run facets

6f02f60

Signed-off-by: wslulciuc <[email protected]>

Remove query for facets in LineageDao.getCurrentRuns()

f6039c6

Signed-off-by: wslulciuc <[email protected]>

Add class RunData

5b565d7

Signed-off-by: wslulciuc <[email protected]>

Limit inserts on spark_unknown and spark.logicalPlan facets

34d9a5a

Signed-off-by: wslulciuc <[email protected]>

rebase commit and fix it to pass tests

fc1754b

Signed-off-by: Pawel Leszczynski <[email protected]>

pawel-big-lebowski force-pushed the feature/ol-facet-tables branch 3 times, most recently from 76e62b0 to 0322664 Compare January 2, 2023 08:20

pawel-big-lebowski marked this pull request as ready for review January 2, 2023 09:04

facet migration with tests

ad0fa19

Signed-off-by: Pawel Leszczynski <[email protected]>

pawel-big-lebowski force-pushed the feature/ol-facet-tables branch from 0322664 to ad0fa19 Compare January 2, 2023 09:46

pawel-big-lebowski requested a review from mobuchowski January 4, 2023 16:03

collado-mike reviewed Jan 5, 2023

View reviewed changes

wslulciuc commented Jan 5, 2023

View reviewed changes

pawel-big-lebowski mentioned this pull request Jan 9, 2023

OL facets - PR1 - create & write new events to new tables while not reading them #2350

Merged

7 tasks

pawel-big-lebowski marked this pull request as draft January 17, 2023 09:38

pawel-big-lebowski closed this Jan 17, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add OL facet tables #2152

Add OL facet tables #2152

wslulciuc commented Sep 30, 2022

pawel-big-lebowski left a comment

pawel-big-lebowski Dec 5, 2022

codecov bot commented Dec 5, 2022 •

edited

Loading

mobuchowski commented Dec 21, 2022

mobuchowski Dec 21, 2022

mobuchowski Dec 21, 2022

wslulciuc Jan 4, 2023 •

edited

Loading

pawel-big-lebowski Jan 4, 2023

collado-mike Jan 5, 2023

pawel-big-lebowski commented Jan 2, 2023

wslulciuc commented Jan 4, 2023

collado-mike Jan 5, 2023

pawel-big-lebowski Jan 10, 2023

collado-mike Jan 5, 2023

wslulciuc Jan 5, 2023

pawel-big-lebowski Jan 10, 2023

collado-mike Jan 5, 2023

pawel-big-lebowski commented Jan 5, 2023

wslulciuc Jan 5, 2023

wslulciuc Jan 5, 2023

pawel-big-lebowski Jan 13, 2023

wslulciuc Jan 5, 2023

wslulciuc Jan 5, 2023

wslulciuc Jan 5, 2023

wslulciuc Jan 5, 2023

wslulciuc Jan 5, 2023

pawel-big-lebowski Jan 5, 2023

wslulciuc Jan 5, 2023 •

edited

Loading

wslulciuc Jan 5, 2023

wslulciuc commented Jan 5, 2023 •

edited

Loading

pawel-big-lebowski commented Jan 17, 2023


		@Setter private Integer chunkSize = null;

		@Setter private boolean triggeredByCommand = false;

Add OL facet tables #2152

Add OL facet tables #2152

Conversation

wslulciuc commented Sep 30, 2022

Problem

Solution

Checklist

pawel-big-lebowski left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

codecov bot commented Dec 5, 2022 • edited Loading

Codecov Report

mobuchowski commented Dec 21, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wslulciuc Jan 4, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

pawel-big-lebowski commented Jan 2, 2023

wslulciuc commented Jan 4, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

pawel-big-lebowski commented Jan 5, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wslulciuc Jan 5, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wslulciuc commented Jan 5, 2023 • edited Loading

pawel-big-lebowski commented Jan 17, 2023

codecov bot commented Dec 5, 2022 •

edited

Loading

wslulciuc Jan 4, 2023 •

edited

Loading

wslulciuc Jan 5, 2023 •

edited

Loading

wslulciuc commented Jan 5, 2023 •

edited

Loading