Jan 29 2018
06:20 AM
- last edited on
Apr 07 2022
04:52 PM
by
TechCommunityAP
Jan 29 2018
06:20 AM
- last edited on
Apr 07 2022
04:52 PM
by
TechCommunityAP
I have configured custom OMS workspace with the DNS Analytics solution enabled. For testing purposes I have generated 43K+ random name resolution queries on test VM against the DNS server that is reporting to OMS. When searching collected data of the DnsEvents type, I see that only handful of queries originating from the test VM are captured on the OMS side. There is great discrepancy in what I expect to be captured and what I see in Log Analytics search results.
Jan 29 2018 07:25 AM
Hi,
Possible reasons could be:
1. Ingestion time - it usually takes around 10 minutes for events to be ingested and searchable)
2. Client capping - the UI client caps the results at 10K. The API will return the full set of results.
3. Query - by default, queries are not sorted by anything. When you review your query results, bare that in mind.
Noa
Jan 29 2018 07:32 AM
Noa,
I ran Log Analytics searches few hours after generating random name resolution lookup queries so data should have been ingested by then.
Besides that, the number of results is far from the UI limit.
Vedran
Jan 30 2018 06:27 AM
Hi Verdan, in that case I would contact support to review what went wrong.
Noa
Jan 30 2018 06:30 AM
SolutionNoa,
I figured it out in the meantime. Random domain name lookup queries were generated with characters which are invalid according to the DNS RFC specification. After excluding invalid characters from lookups, I am getting results which are aligned with the testing scenario.
Kind regards,
Vedran
Feb 05 2018 06:00 AM
Feb 05 2018 07:10 AM
Glad to help, but actually there is nothing to share. I didn't come up with particularly special Log Analytics Query Language query.
I made amendments on the DNS client side (device generating random DNS queries for testing purposes), meaning that random DNS queries were modified to exclude characters which are invalid per DNS RFC.