Child pages
  • Terracotta DB 10.2 Release Notes

Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  • AWS as a supported platform for Terracotta DB
  • Terracotta DB now supports docker-based container orchestration using Kubernetes
  • Advanced support for resilient failover through improved "Failover-tuning" option
  • Resiliency support for improved re-connection for both Ehcache and TCStore clients
  • Security enhancements including SSL support for Encryption and Authentication
  • SQL-like query capability using new Terracotta Textual Query Language 
  • Ad Hoc Query capability within Terracotta Management Console with new Terracotta DB Explorer
  • Numerous operational usability enhancements including Latency statistics and Query statistics monitoring

Summary of Changes 10.2.0.4

Resolved

  • 3891 – 10.2 client is now able to connect to 10.3 server
  • 3834 – Enabled voter to be able to connect to secure clusters

Known Issues

  • 3119 - Cluster tool fails to connect to server when the bind address is different from host

Summary of Changes 10.2.0.3

Resolved

  • 3417 – Corrected the exception raised when failing to create a TCStore index due to insufficient memory availability.
  • 3290 – Resolved issue where TMC may incorrectly display ACTIVE state for a Terracotta DB server which is shutdown.

Known Issues

  • 3119 - Cluster tool fails to connect to server when the bind address is different from host

Summary of Changes 10.2.0.2

Resolved

  • 3065 – Resolved NPE in FRS when restarting standalone TCStore
  • 3066 - Resolved hybrid storage mode race condition that may occur between appender and reader threads for doing fsync when there are read and writes in parallel; which then may lead to server crash.
  • 3152 - The security documentation updated with information on certificate rotation

Known Issues

  • 3119 - Cluster tool fails to connect to server when the bind address is different from host

Summary of Changes 10.2.0.1

Resolved

  • 2450 - Two servers in a stripe can be mistakenly displayed as ACTIVE in TMC

  • 2801 - "common name" in the server’s TLS/SSL certificate is not validated properly

  • 2966 - Cluster tool "status" command shows inconsistent stripe contents

  • 2969 - ipwhitelist-reload returns success even if the command fails

  • 2993 - Improve impact of thread interrupt during reconnection

  • 3013/3138 – Increase the default of cluster-tool connection to server timeout and added option to control this timeout in cluster-tool command line
  • 3014 - fixed multiple cases where reconnect can fail

  • 3022 - Reconnect attempts should use a distinct timeout

...