Commits

Branch selector

 master

User selector

aka-netzero

All time

Commit History

Commits on May 4, 2020

• Reduce log noise when no checkpoints were encountered

  aka-netzero

  committed
  02563ac

  Copy full SHA for 02563ac

Commits on May 1, 2020

• Tying the registration/publication of coordinator delay metric to whether or not the replicator instance has leadership.

  aka-netzero

  committed
  d75fc7c

  Copy full SHA for d75fc7c

Commits on Mar 16, 2020

• Remove commented lines

  aka-netzero

  committed
  5c44907

  Copy full SHA for 5c44907

Commits on Mar 3, 2020

• Dropping the default binlog size down, it *shouldn't* have any impact on the existing tests, but is necessary for the fallback test (which requires that we are able to purge binlog files which aren…

  Show description for 80acd48
  aka-netzero

  committed
  80acd48

  Copy full SHA for 80acd48

• Removing all copy/paste instances of the runMysqlScript sub and accompanying subs to enable it, and replacing them

  Show description for 21f5da4
  aka-netzero

  committed
  21f5da4

  Copy full SHA for 21f5da4

• Add default values into the MySQLConfiguration class. These defaults (or some variation thereof) are present

  Show description for a864ded
  aka-netzero

  committed
  a864ded

  Copy full SHA for a864ded

• Adding a configuration variable and supporting code to set GtidSetFallbackToPurged to true/false in BinaryLogSupplier. This

  Show description for 7254f0a
  aka-netzero

  committed
  7254f0a

  Copy full SHA for 7254f0a

Commits on Feb 7, 2020

• Explicitly flush per table mutations in the HBase applier, this will prevent background buffering of the mutations.

  Show description for 7932192
  aka-netzero

  committed
  7932192

  Copy full SHA for 7932192

Commits on Feb 5, 2020

• Increasing the amount of information written in the event of an error deserealizing a row.

  aka-netzero

  committed
  ca6e65a

  Copy full SHA for ca6e65a

• Adding a try/catch around deserialization of individual cells in order to increase the amount of information we output in the event of an error.

  aka-netzero

  committed
  144e933

  Copy full SHA for 144e933

• Moving the check to ensure that the incoming event contains the same number of columns as the schema cache up a level, to ensure that all event types that require deserialization are checked.

  aka-netzero

  committed
  0571a8e

  Copy full SHA for 0571a8e

Commits on Nov 4, 2019

• Patching TRID Partition method to avoid an issue whereby Math.abs() returns a negative number in the circumstance that event.getHeader().getEventTransaction().getIdentifier().hashCode() return Inte…

  Show description for 904b4c0
  aka-netzero

  committed
  904b4c0

  Copy full SHA for 904b4c0

Commits on Oct 8, 2019

• Removing leftover debug message

  aka-netzero

  committed
  baa55ca

  Copy full SHA for baa55ca

• Mistakenly duplicated the check for MAX rather than MAX && MIN

  aka-netzero

  committed
  a08db05

  Copy full SHA for a08db05

• Added support for augment.context.exclude.pattern configuration option, this allows for support of regex based excludes. This was a feature that existed in 014

  aka-netzero

  committed
  ac25a06

  Copy full SHA for ac25a06

Commits on Oct 7, 2019

• Clearing the timeout so that it will be randomised in the next iteration

  aka-netzero

  committed
  4ed0829

  Copy full SHA for 4ed0829

• Updating the version of the bigtable-hbase-1.x package to the latest one we have available

  aka-netzero

  committed
  a2b62c1

  Copy full SHA for a2b62c1

Commits on Oct 2, 2019

• Introduced an option to randomize the timeout length for HBase buffers. This is intended to test if the post-startup behaviour we have observed in the past, in which our writes to BigTable occur in…

  Show description for e67feea
  aka-netzero

  committed
  e67feea

  Copy full SHA for e67feea

Commits on Aug 27, 2019

• First pass at converting transaction_uuid to GTID

  aka-netzero

  committed
  21882fd

  Copy full SHA for 21882fd

Commits on Jul 31, 2019

• Comment out the unit tests for DATE,DATETIME and TIME while we rework them. This is mainly to allow the CI builds to move past these tests for the time being

  aka-netzero

  committed
  0633e9c

  Copy full SHA for 0633e9c

• Added a perTestConfiguration method to the ReplicatorHBasePipelineIntegrationTest class. Combined with the changes to the TestRunner class this allows tests to contain contradictory configurations …

  Show description for 5ec2391
  aka-netzero

  committed
  5ec2391

  Copy full SHA for 5ec2391

• Removing the configuration option to convert DATE and TIMESTAMPS to LONG. It would be ideal to have *ONLY* TIMESTAMPS returned as LONG, but the fact that this configuration option also modifies the…

  Show description for b0f5d43
  aka-netzero

  committed
  b0f5d43

  Copy full SHA for b0f5d43

Commits on Jul 30, 2019

• Set the MESSAGE_FORMAT to JSON to prevent errors regarding not having a SCHEMA_REGISTRY URL

  aka-netzero

  committed
  fd7791d

  Copy full SHA for fd7791d

• value_before was not present, but value itself is. This fixes an issue in which a DELETE would cause an exception to be thrown complaining about not having support for tables without primary keys

  aka-netzero

  committed
  f054135

  Copy full SHA for f054135

Commits on Jul 18, 2019

• Fixed incorrect gtidSet in CHECKPOINT_DEFAULT, which was preventing the binlog connector from successfully connecting to MySQL. Additionally re-enabled the full suite of HBase tests now that everyt…

  Show description for d56f2a6
  aka-netzero

  committed
  d56f2a6

  Copy full SHA for d56f2a6

Commits on Jul 17, 2019

• Trying to resolve the Travis CI issues, according to some Googling this *should* fix the problem

  aka-netzero

  committed
  860a934

  Copy full SHA for 860a934

Commits on Jun 4, 2019

• Re-upped the per-thread flush buffer size limit, this will allow us to more quickly recover from backups. To be moved to configuration file in future commit

  aka-netzero

  committed
  c44dc25

  Copy full SHA for c44dc25

Commits on Apr 4, 2019

• Changing all exceptions to be runtime exceptions, in production we will be running two instances of each replicator to allow for failover, so don't bother catching exceptions just to potentially ge…

  Show description for 319ff8e
  aka-netzero

  committed
  319ff8e

  Copy full SHA for 319ff8e

• Fixing namespacing issue for metric

  aka-netzero

  committed
  3707548

  Copy full SHA for 3707548

Commits on Mar 21, 2019

• Throw a runtime exception if any of the writes to BigTable/HBase fail, this will prevent data loss

  aka-netzero

  committed
  ff514a5

  Copy full SHA for ff514a5

Commits on Mar 18, 2019

• Modifying existing behaviour to support position override when checkpoint type is GTID. This allows a less cumbersome bootstrapping as it allows you to specify where to begin, but then allows the c…

  Show description for b3e945b
  aka-netzero

  committed
  b3e945b

  Copy full SHA for b3e945b

• Lowercase table names when writing to hbase/bigtable. All HBase tables are already lowercase, so lets not change a good thing

  aka-netzero

  committed
  23114b0

  Copy full SHA for 23114b0

• Upping the default FLUSH_BUFFER_SIZE. This was modified during testing and when left this low will ultimately cause the replicator to die due to being unable to keep up with increasing queue sizes

  aka-netzero

  committed
  5dbca77

  Copy full SHA for 5dbca77

Commits on Feb 27, 2019

• Remove two unnecessary messages and upgrade the level of logging for push timeouts

  aka-netzero

  committed
  b392cc5

  Copy full SHA for b392cc5

• Removed currentTransactionSchemaName variable and the logic surrounding it to prevent logging of non-replicated statements

  aka-netzero

  committed
  6926b5a

  Copy full SHA for 6926b5a