OK…so I have a larger install, storing about 1500 events a second. We have our indices set to 1 day and purge after 30 count. This is a fresh build of Graylog on Ubuntu 16.04, 24 GB ram, 8 cores of proc, 6 TB of storage, running Graylog 2.3.1
We are seeing this error in the logs, but I’m not sure what to do about it.
**{"type":"illegal_argument_exception","reason":"Limit of total fields [1000] in index [graylog_82] has been exceeded"}**
The Windows EventLog sends structured messages, so if you have many different events in the Windows EventLog with vastly different field names and don’t consolidate these fields in Graylog (e. g. via the processing pipelines), you can get 1000+ different field names.