Is SysID Conflict a real thing?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
yesterday - last edited yesterday
Hi Community,
I am working on a merger project in which we need to consolidate foundational data to support CMDB consolidation across two independent production instances.
My concern is about sys_id conflicts, since both instances have been running independently for years and there is a real chance the same sys_id exists on both sides, pointing to completely different records.
Has anyone dealt with this before? Specifically:
- Is sys_id conflict a real risk when migrating foundational data via IDR or Import XML?
- How did you approach identifying and resolving conflicts at scale?
- What mechanism did you use for CMDB/foundational data merger and why?
Any real-world experience or recommendations would be greatly appreciated!
Thanks
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
yesterday
Hi @vishnukelam
I worked on couple of migration project ,But there all were fresh instances.
What I can forsee.
1. During Migrating your foundational data using Xml export/import, can cause duplicate sys_id.
When two separate instances (e.g., Instance A and Instance B) generate records independently, it is statistically possible to generate the same 32-character sys_id for entirely different records.
e.g group record. I will suggest before importing you must first do a full "collision check" to ensure the sys_id does not already exist in the target for a different record.
2. If you use using transform map with coalescing on unique field, generated record's sys_id would be unique. The safest method is to not bring over the sys_id during the import. Instead, map based on a unique business key

