v1.18.0 (02/05/2019)

Enhancements:

• #2230 ISL and switch details can be open from a flow path on GUI (Issue: #1740)

Bug Fixes:

• #2245 Add error response for sync meters on umetered flow (Issue: #2043)

Improvements:

• #2002 Introduce the data model with flow encapsulation, flow paths and resource pools. (Issue: #1519)
• #2103 Adopt the new data model via a wrapper entity (Issue: #2018)
• #2106 Implement transactional resource pools (Issue: #2017)
• #2138 Migration scripts for the new data model. (Issue: #1519)
• #2247 Add core properties auto create metrics,tagks,tagvs to opentsdb.conf
• #2301 Re-define OGM mapping for Flow, FlowPath and PathSegments to improve DAO performance

Other changes:

• #1906 Protected paths design [area/docs] (Issue: #1232)
• #1904 Design for flow rerouting using H&S approach [area/docs]

For the complete list of changes, check out the commit log.

Upgrade notes

If you have an older version of Kilda installed, then you must migrate the data stored in Neo4j
before you deploy and start this version.

The data migration steps:

• If any of Kilda components is running, shut it down. Neo4j must be up and a backup is created.
• Execute 1-update-constraints-changelog.xml
  and 2-migration-changelog.xml scripts using Liquigraph tools
  (see migration README.md).
• Deploy and start all Kilda components of this version.
• After CRUD tests are passed and the system is being operational with no issues for some period, you need to execute the cleanup script
  3-cleanup-changelog.xml.

In case of any issue you can rollback the migration:

• Execute queries from rollback-migration.cql
  to revert changes made by 2-migration-changelog.xml.
• Execute queries from rollback-constaints.cql
  to revert changes made by 1-update-constraints-changelog.xml.

IMPORTANT: The changes made by the cleanup script (3-cleanup-changelog.xml) can't be reverted!