…ider

Summary:
Add access key block to request body only when the fields are present and have valid inputs

Original commit: 839f201 / D52791

Test Plan: Manually test

Reviewers: amalyshev, nsingh

Reviewed By: nsingh

Subscribers: yugaware

Differential Revision: https://phorge.dev.yugabyte.com/D52843

…ash -c drops --input in yb_platform_backup.sh

Summary:
Original diff: https://phorge.dev.yugabyte.com/D52820 (d6eb93e)

 Shell script expansion issue when the array is a part of a longer quoted string.

```
    kubectl -n "${k8s_namespace}" exec -it "${k8s_pod}" -c yugaware -- /bin/bash -c \
            "${backup_script} restore ${restore_args[@]}"
```

got expanded to

```
    kubectl -n yw-ns-jenk-id19549-7b5a150ce1-secondary exec -it yw-ns-jenk-id19549-7b5a150ce1-secondary-yugaware-0 -c yugaware -- /bin/bash -c '/opt/yugabyte/devops/bin/yb_platform_backup.sh restore -v' --input /opt/yugabyte/backup_26-05-06-14-27.tgz --skip_old_files
```

Test Plan:
itest run has confirmed (passed) it.

https://jenkins.dev.yugabyte.com/job/dev-itest-job/35381/artifact/logs/2.13.3.0_testk8sybabackup-k8s-rf3_20260507_085030/testk8sybabackup-k8s-rf3fvbijkhy.log

```

Unable to use a TTY - input is not a terminal or the right kind of file
+ backup_file=backup_26-05-07-09-10.tgz
+ backup_file=backup_26-05-07-09-10.tgz
+ verbose_flag=
+ [[ true == true ]]
+ verbose_flag=-v
+ backup_script=/opt/yugabyte/devops/bin/yb_platform_backup.sh
+ restore_args=(${verbose_flag} --input ${K8S_BACKUP_DIR}/${backup_file} --skip_old_files)
+ [[ false = true ]]
+ [[ false = true ]]
+ cmd=("$backup_script" restore "${restore_args[@]}")
++ printf '%q ' /opt/yugabyte/devops/bin/yb_platform_backup.sh restore -v --input /opt/yugabyte/backup_26-05-07-09-10.tgz --skip_old_files
+ kubectl -n yw-ns-jenk-ci35381-64734e5830-secondary exec -it yw-ns-jenk-ci35381-64734e5830-secondary-yugaware-0 -c yugaware -- /bin/bash -c '/opt/yugabyte/devops/bin/yb_platform_backup.sh restore -v --input /opt/yugabyte/backup_26-05-07-09-10.tgz --skip_old_files '

```

Reviewers: anijhawan, nbhatia, ayush.kushwaha

Reviewed By: ayush.kushwaha

Subscribers: nikhil, yugaware, ayush.kushwaha

Differential Revision: https://phorge.dev.yugabyte.com/D52902

…ong time (> 2 hours) even in 3 node clusters

Summary:
Disabled the redaction for postgres sql logs , as they were causing high cpu usage also
taking lot of time for support bundle creation.
Original Commit : fe9ec80 / D52857

Test Plan:
Manual Testing

Dev Testing Steps:

Setup:

Universe with 3 db nodes
Generated lot of postgres sql logs under ~/tserver/logs for each node
total of 6000+  postgres sql files including all nodes
each file < 5 MB in size

Support Bundles Created

Only Universe Logs:
Complete Support Bundle except prometheus
Time range - last 3 days

Observation:

Before Fix:

Time taken more than 20 min , stuck at 38% for Complete Support Bundle except prometheus  => manually aborted
Same observation for only Universe logs => Manually Aborted
CPU usage => 300%

After Fix :

Time taken more than < 3 min
Complete Support Bundle : 296 MB
Only Universe Logs : 294 MB

CPU usage => 120%

Reviewers: vkumar, vbansal, aaryan.chauhan, anijhawan

Reviewed By: vbansal, aaryan.chauhan

Subscribers: yugaware, agunta, agarg

Differential Revision: https://phorge.dev.yugabyte.com/D52876

…_lsn is greater than last_seen_lsn

Summary:
#### Backport Description
No merge conflicts were encountered.

#### Original Description

##### Code changes summary
When the walsender receives a` confirmed_flush_lsn` as an acknowledgement, it sends `restart_lsn`, calculated as `confirmed_flush_lsn + 1` to the virtual WAL. The VWAL has a mechanism to move the restart time forward when there is nothing left to stream. To detect that there is nothing more left to stream, it compares the `restart_lsn` with the `last_seen_lsn` (the last lsn generated for a record by VWAL). If they are equal, restart time is moved forward.

However, since walsender increments `restart_lsn`, the equality condition would never be true. As a result, the mechanism of moving restart time forward did not work. This diff fixes this issue by changing the equality check to a greater than equal to check.

##### Considerations for colocated tables
NA

##### Upgrade and rollback considerations
NA

##### Considerations for connector
NA

Original commit: 9202b63 / D52037

Test Plan:
New test, this brings walsender into the test's scope:
  - ./yb_build.sh --java-test 'org.yb.pgsql.TestPgReplicationSlot#testRestartTimeMovementWithNoWorkload'

Modified an existing test:
  - ./yb_build.sh --cxx-test integration-tests_cdcsdk_consumption_consistent_changes-test --gtest_filter CDCSDKConsumptionConsistentChangesTest.TestMovingRestartTimeForwardWhenNothingToStream

Reviewers: #db-approvers, skumar, asrinivasan, devansh.singhal

Reviewed By: devansh.singhal

Subscribers: svc_phabricator, ycdcxcluster

Tags: #jenkins-ready

Differential Revision: https://phorge.dev.yugabyte.com/D52850

…nd the correct payload for useRoles or cron-based schedules

Summary: Backport of this [[ https://phorge.dev.yugabyte.com/D52693 | diff ]]

Test Plan: Tested manually

Reviewers: lsangappa

Reviewed By: lsangappa

Subscribers: yugaware

Differential Revision: https://phorge.dev.yugabyte.com/D52887