How can I troubleshoot log ingestion issues with a log source?
Follow the tips below while troubleshooting log ingestion issues in Panther:
Confirm whether Panther is receiving any data.
You can check the log source. It should say “Last data received X minutes ago” if you are expecting it to receive logs every few minutes.
If it's a Data Transport source, such as an S3 bucket, check to verify that the source has received data since setting up the ingestion with Panther. See this KB article for more information: 📄 How come no data is coming in for a new S3 log source in Panther?
If there is data received:
The issue could be an error from Panther reading the source (e.g., the log is encrypted and Panther expects it to not be, or the log file is not in the right format. See:📄 How to resolve "Failed to infer schema... error found in byte" when inferring schema in Panther). To check for this, you can run a query in Data Explorer to search for the log source name in the panther_monitor.data_audit table.
The issue could be a misclassification error where Panther successfully reads it but fails to classify. To check for this, you can run a query in Data Explorer to search for the log source name in the panther_monitor.public.classification_failures table.
If there is no data received yet then follow the suggestions below. Note that this could include seeing "Last data received X hours ago" in a situation where you are expecting data every few minutes.
Check whether bucket notifications are set up correctly. If there are no bucket notifications, then Panther will display “No data received yet” next to the log source. See this KB article for more information: 📄 Panther Log Source error: "Bucket notifications are not properly configured"
If you created an IAM role manually for an S3 log source and you are not receiving data, see this article:📄 No data flow or errors after creating IAM role manually for S3 source in Panther
Ensure that your SNS topic has therawmessagedeliveryoption disabled. See the documentation for more information: S3 Source - Create SNS subscription
If you're using multiple schemas in a single log source such as S3 or another custom source, make sure the schemas make adequate use of required fields. Not doing this can result in some logs successfully ingesting with an unexpected schema/log type, which then leaves the expected schema/log type empty.
Please also see these related articles: