Download the release on GitHub.
Features
Section titled “Features”Make the grok parser precise
Section titled “Make the grok parser precise”The grok parser now allows better control over the schema inference.
The grok parser can now be directly used when reading input, allowing for
read grok.
By @IyeOnline in #4657.
Add sample operator
Section titled “Add sample operator”A new sample operator now provides the ability to dynamically sample input
data based on the frequency of the receiving events allowing relative sampling
in situations of varying load.
Bug Fixes
Section titled “Bug Fixes”Switch the index to basic messaging
Section titled “Switch the index to basic messaging”We fixed a bug that sometimes caused the tenzir-node process to hang on
shutdown. This was most likely to happen when the node shut down immediately
after starting up, e.g., because of an invalid configuration file.
Fix double-closing fds in the python operator
Section titled “Fix double-closing fds in the python operator”Fixed a bug in the python operator that could lead to random valid file descriptors in the parent process being closed prematurely.
Enable azure-blob-storage in the DEB and RPM packages
Section titled “Enable azure-blob-storage in the DEB and RPM packages”The azure-blob-storage connector is now also available in the static linux
binary distributions.
Fix incorrect context updates count in lookup metrics
Section titled “Fix incorrect context updates count in lookup metrics”We fixed a bug that caused the context_updates field in metrics lookup to be
reported once per field specified in the corresponding lookup operator instead
of being reported once per operator in total.
By @dominiklohmann in #4655.