Discovery Best (better) Practices

[doug_schulze]

All new for 2023, attached is a zip file of some of the better practices I've come across in the 14+ years of doing discovery deployments and support...Your mileage may vary, there are many things to consider in a Discovery Deployment

1. MidServer host selection
2. MidServer application management
3. Use of Credentials
4. Management of schedules
5. Utilizing behaviors
6. Cluster groups
7. Probes/sensors/Patterns
8. Properties
9. Troubleshooting

DiscoveryBetterPractices.zip

[vinay_jbpt]

It is very useful thank you for sharing...

[MarkH7243205430]

Hey Doug, appreciate the refresh of the docs. Just checking to see if that is a typo in the BPD01 doc for MID server sizing saying:

1. 40Gb RAM

[piotrkopytynski]

Hello @doug_schulze 

We have a MID server in GCP that we use for IP-based discovery. Looking at the MID server logs we see a significant number of requests (1269 API calls within 5 hours for a typical MID server) against our instance to /api/now/pattern_library/GetNonCITypeAttributes/{table_names} (/nav_to.do?uri=sys_ws_operation.do?sys_id=310fd110df153200d7dc97a83df263b6). I have opened a case with support about that number but it looks like it is working as designed. But why that content cannot be cached on the MID server? Why do we have to make so many API calls to fetch that information?

[Thomas Powe]

@piotrkopytynski  I do believe this is due to the fact that MID Server are more like pass-through, and do not hold any information.  They only hold binaries and Discovery patterns.

[Ashu Rohtak]

I am very late person who missed to download this zip file 😞. @doug_schulze can you please re-upload the "best practice" document so people like me can grab this opportunity again.🙂

TIA