Update dependency io.camunda:zeebe-client-java to v8.5.7

[renovate]

This PR contains the following updates:

Package	Change	Age	Adoption	Passing	Confidence
io.camunda:zeebe-client-java (source)	8.5.3 -> 8.5.7

---

Release Notes

camunda/zeebe (io.camunda:zeebe-client-java)

v8.5.7

Compare Source

Bug Fixes

Broker

• Exporter arguments are always converted to maps instead of list (#​4552)

Misc

• Multi instance body finishes with less inner activities than expected (#​20958)
• NPE in BpmnStateBehavior.getParentElementInstanceContext "parentElementInstance" is null (#​20762)
• Possible memory corruption leading to SIGSEGV (#​20555)
• Exporting decision evaluation records breaks with large evaluationFailureMessage values (#​20248)
• CI: Zeebe build fails due to SBE build errors (#​16029)
• NullPointerException in ActivateProcessInstanceBatchProcessor.createChildInstanceRecord (#​12918)

Maintenance

• Renovate updates project parent poms but shouldn't (#​20856)
• Refactor fold method argument order in Either interface (#​20588)
• Sequencer#isEntryValid is slow because it has to calculate the full record size (#​19225)

Merged Pull Requests

• build: fix fetching the last push commit (#​21861)
• Allow instantiating exporter configuration with list fields/members (#​21847)
• fix: make SBE code generation thread safe (#​21801)
• [Backport stable/8.4] build: replace protolock with buf CLI (#​21773)
• build: add buf configuration file (#​21696)
• build: replace protolock with buf CLI (#​21616)
• fix: prevent completing multi-instance body until the completion/termination of all its children (#​21442)
• bugfix: set a limit to evaluationFailureMessage size (#​21320)
• refactor: record value and metadata length is always non-zero (#​21311)
• Unlearn spotless to ratchet from main (#​21200)
• feat: allow dns resolution to fall back to TCP (#​21066)
• test: ensure that controlled actor scheduler starts idle (#​21043)
• refactor: Change order of fold method arguments in Either interface (#​20889)
• fix: resetting a reader does not invalidate returned entries (#​20720)
• fix: run scheduled tasks at a given timestamp (#​18986)

v8.5.6

Compare Source

version:8.5.6

Enhancements

Misc

• Add Intermediate None Event to StraightThroughProcessingLoopValidator (#​20308)

Bug Fixes

Broker

• Limit the number of element in the process caches (#​15533)

Gateway

• Considerable amount of GRPC error details can trigger ingress response header limit (#​11284)

Java Client

• Job worker should apply backoff on job polling in case of service unavailable (#​4443)

Go Client

• Job worker should apply backoff on job polling in case of service unavailable (#​4443)

Misc

• StackOverflowError during process instance modification (#​20307)
• Flaky test shouldBeThreadSafe: io.camunda.zeebe.client.impl.oauth.OAuthCredentialsCacheTest (#​20136)
• Snapshot detected as corrupted because of incorrect checksum comparison (#​19984)
• No executable process is found after transformation (#​19834)
• ZEEBE_GATEWAY_SECURITY_AUTHENTICATION_MODE:none does not disable Identity-based authentication (#​18726)
• java.lang.IllegalArgumentException: type must not be empty (#​18139)
• Oauth token refresh can trigger rate limiting and log errors (#​13832)

Maintenance

• Strange use of ObjectMapper in ES/OS DTO (#​20556)
• Snyk scan for Zeebe is broken after moving modules around (#​20011)
• gRPC thread pool saturated by blocked Identity calls (#​18697)

Merged Pull Requests

• Fix Snyk scanning action for Zeebe (#​20610)
• Remove grpc thread config (#​20591)
• Dont use object mapper in DTO to ES / OS (#​20560)
• Reset RecordingExporter before each ZeebeIntegration test setup (#​20524)
• Create incident when type expression is null or empty (#​20517)
• Fix OauthCredentialsCache concurrency issues (#​20429)
• Terminate elements iteratively instead of recursively during PI modification (#​20425)
• Test for snapshot corruption when restoring from backup. (#​20415)
• Add logging to setup commands (#​20389)
• test: fix potential race condition (#​20334)
• fix: java client reminder action (#​20322)
• Add intermediate throw event StraightThroughProcessingLoopValidator (#​20316)
• fix: ensure controlled actor scheduler is closed after tests (#​20296)
• test: prevent db transaction corruption by test (#​20288)
• Fix flaky MessageStreamProcessorTest (#​20286)
• ci: add java client reminder workflow (#​20285)
• test: fix flaky raft heartbeat test (#​20079)
• test: snapshot false negative corruption unit test (#​20040)
• fix: Allow single deployment of multiple BPMN processes with the same resource name (#​20026)
• feat: allow overriding default maxInboundMetadataSize through config and change default to 16KB (#​20002)
• fix: do not use combined checksum for snapshot integrity (#​19988)
• deps: Update dependency io.grpc:grpc-bom to v1.65.0 (main) (#​19858)
• Allow disabling REST gateway security through configuration (#​19823)
• fix: limit validation result output to a certain size and print the number of remaining errors (#​19600)
• change grpc executor to virtual threads (#​19005)
• feat: implementing bounded cache (#​17581)

v8.5.5

Compare Source

version:8.5.5

Bug Fixes

Misc

• Snapshot detected as corrupted because of incorrect checksum comparison (#​19984)
• Failed to close snapshot store (#​17452)

Merged Pull Requests

• fix: apply correct license header for 8.5 (#​20008)
• Backport build-platform-docker GHA (#​19740)
• fix: do not close snapshot store in RaftContext (#​19719)

v8.5.4

Compare Source

8.5.4

⚠️ There is a bug in Zeebe 8.5.4 which will incorrectly flag a snapshot as corrupted even when it's not. This can cause nodes to fail to start up, which can eventually lead to unavailability. The fix is to update to 8.5.5, which corrects this. Note that there is no data loss or actual corruption, it's a false positive, but it has an impact on the availability of the system.

Bug Fixes

Misc

• Spring SDK 8.5.3 requires JDK 21+ (#​19658)
• Invalid offset when reading record from an older version (#​19379)
• Cannot take backup to Azure when using managed identity (#​19104)
• Default tenant id not picked up by Spring Zeebe SDK @​JobWorker annotation (#​18975)

Merged Pull Requests

• Backport build-platform-docker GHA (#​19740)
• fix: check version before reading metadata length (#​19582)
• test: wait until deployment is distributed (#​19466)
• build: unpin msal4j version (#​19463)
• Stabilize ReplayStateRandomizedPropertyTest (#​19393)
• fix: remove static modifier of TestStandaloneBroker in client IT test classes (#​19362)
• fix flaky test DeploymentClusteredTest.shouldRedistributeDeploymentWhenDeploymentPartitionIsRestarted (#​19358)
• Fixing default tenant IDs not being picked up from properties (#​19299)
• Avoid NPE when parameter names are not available (#​19082)
• Checksums calculated with full file checksums (#​18730)
• feat: removed combined checksum check (#​18722)
• feat: enable full file checksum (#​18721)

---

Configuration

📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - "after 10pm every weekday,before 5am every weekday,every weekend" (UTC).

🚦 Automerge: Enabled.

♻ Rebasing: Whenever PR is behind base branch, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.