Vulnerability response - Reports based on services
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
09-15-2026 06:30 AM
Hi all.
We have implemented Vulnerability response for servers and the process as a whole is working great. However, from the reporting perspective, we are looking to get a little than what I think is offered out of box. Leadership would like to see by "application service" how many vulnerabilities each service has. This would allow them to figure out which of our most critical applications have the most vulnerabilities and allow additional filtering.
From what I can tell there are only a couple ways to accomplish this...
1: CMDB Query builder to generate the data on a monthly basis - not ideal as it would be a huge data set.
2: Database view - Not ideal as the data loads every time you view it.
3: Custom field to the VIT: This would let us capture the service(s) related to the VITs CI
Has anyone else had to achieve something similar? if so, how did you achieve it?
Thanks
Josh
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
2 weeks ago
Hi Josh,
Your option 3 is the right instinct, but with one important correction: do not maintain the service on the Vulnerable Item by hand, and do not use a single-value field for it.
The service relationship already exists in the CMDB - the Vulnerable Item points to a CI, and that CI is related to its application service(s) through cmdb_rel_ci (Service Mapping / application service relationships). That relationship should stay the source of truth. If the service is copied onto the VIT manually, it will drift the moment a CI moves between services.
The pattern that generally works best:
1. For ad-hoc or monthly reporting, report by traversal: Vulnerable Item -> CI -> related application service. A database view over the Vulnerable Item, cmdb_ci and cmdb_rel_ci (filtered to the service relationship type) gives you one row per VIT-service pair, which is exactly what you need for counting by service. CMDB Query Builder can do the same exploration, but treat it as the design/prototype step, not the monthly production report.
2. If your reporting tool cannot traverse that chain, or the database view is too slow at your volume, add a derived field on the VIT - but populate it automatically, never by hand. A scheduled job, or a business rule on VIT insert/update (and on CI relationship change), can read the CI's related services and write them to a list-type field on the VIT. The list type matters: a CI can belong to more than one application service, and a single reference field will silently drop the others.
3. Keep a "no service" bucket in the report. CIs with no related application service will otherwise vanish from the counts, and leadership will read that as zero risk. The unmapped population is usually the most useful finding in the first version of this report.
So: report from the relationship where you can, snapshot it automatically where you cannot, and never let a manually maintained service field become the source of truth.