Hello,
We are a system integrator in Korea building data center EPMS/BMS/DCIM systems on Ignition. Our sites run a Redundant Gateway Pair (Master/Backup), and our customers increasingly require the database layer to be redundant as well, with no data loss.
Environment
- Ignition version: [8.x.x]
- Database: [MS SQL Server / PostgreSQL / MySQL] [version]
- Architecture: Redundant Gateway Pair, [N] tags, history rate s
- Data stored in DB: Tag History, Alarm Journal, Audit Log, Transaction Groups, Script/Named Query writes
Our current understanding (please correct us if anything is wrong)
- Ignition does not natively replicate or synchronize data between databases; DB HA is expected to be handled by the RDBMS.
- Failover Datasource (STANDARD/STICKY) only reroutes queries to a secondary connection and does not resync data back to the primary after recovery.
- Tag History Splitter duplicates Tag History writes only; Transaction Groups, scripts, etc. are not covered.
- Store & Forward buffers history data while the DB is unavailable.
Questions
- Is native database redundancy (replication or automatic resync between primary and secondary DBs) on the roadmap for 8.3 or later?
- For a Redundant Gateway Pair in a mission-critical data center, what DB HA architecture does IA recommend (DB listener only / Failover Datasource / History Splitter / a combination)? Field experience would be greatly appreciated.
- When using Failover Datasource or History Splitter, how should we handle data written only to the secondary DB during an outage (gaps or duplicates after fail-back)? Is there a recommended resync method?
- How should non-history data (Transaction Groups, Alarm Journal, Audit Log, Script/Named Query writes) be protected in this architecture?
Thank you.