Hi
I am new to OT, and I am struggling with a use case that I could really use some advice on, please
I have a test case where we need to send 3.3K logs per second over HEC to Splunk using.
The data is being currently sent into one index and one source type (exporter below)
The data is 3 different sourcetype defined by the event “logs.type” The issue is the speed in searching
The SPL I am using is below, but it is not fast for high volumes
index="murex_logs"
| regex mx.env="dell967srv:15017" ```We have multiple environment sending in the data ```
| regex log.type="http" ```http is one of 3 that the data source could be```
To me this will be very slow 1st running regex is probably slow and also it has to sort out http from the other data (3 types).
I am talking to dev to see if they can send the data on three different exporters (Below)
However I still must run a regex mx.env="dell967srv:15017 to find the environment that I need.
exporters:
splunk_hec/logs_1: # pushed to splunk
token: "a04daf32-68b9-48b2-88a0-6ac53b3ec002"
endpoint:
https://mx33456vm:8088/services/collector source: "mx"
sourcetype: "otel"
index: "murex_logs"
tls:
insecure_skip_verify: true Some of the possible answers I am looking into are
- Use transforms: To create 3 different sourcetype If possible? But can I do this on event data (regex mx.env="dell967srv:15017" )[Purple below], I know I can do it on raw data but not sure about event data (Green below)
- Ask the dev team to send the data by log.type and don’t put all the data into one index (But I will still have to use regex for the host)
This is my log data. The green data is the Raw data – the Purple is the Event data.
Any help would be amazing – thanks in advance
