Getting Data In

Duplicate Siebel log events from Universal Forwarders

Path Finder

Hi we are getting duplicates on log events

Events are :

- multiline

- large (to very large)

- also the files can grow to very large

- open for long time



- when oneshotting the file > no duplicates

- duplicates may arrive on different indexers, or same indexer

- duplicates may arrive in short time span or even up to an hour

- we get them both from Solaris and Redhat hosts

we have played with useAck, increasing filedescriptors, use of parallel mode, increasing bandwidth to try and rule out it being the flow towards the HF.

From logs we do see files being read entirely up to 3 times, but this is not the case for all, so there may be several root causes.

I am now thinking in the direction of writebuffers, but am not sure because I do also see smaller events being duplicated.

we are using these settings for props

EVENT_BREAKER = ([\r\n]+)(?:\w+\s+\w+\s+\d{1})


TRUNCATE = 99999    # because of large messages   
TIME_PREFIX = ^(?:[^\t]+?\t){4} # Format is 2021-01-18 16:03:27.118885 so 26 characters
TIME_FORMAT = %Y-%m-%d %H:%M:%S.%6N
LINE_BREAKER = ([\r\n]+)(?:\w+\s+\w+\s+\d{1}) # newline characters can be part of the message

any pointers from your experience on how to further troubleshoot, or just extra pair of eyes/brain is more than appreciated.



0 Karma
Get Updates on the Splunk Community!

Observability | How to Think About Instrumentation Overhead (White Paper)

Novice observability practitioners are often overly obsessed with performance. They might approach ...

Cloud Platform | Get Resiliency in the Cloud Event (Register Now!)

IDC Report: Enterprises Gain Higher Efficiency and Resiliency With Migration to Cloud  Today many enterprises ...

The Great Resilience Quest: 10th Leaderboard Update

The tenth leaderboard update (11.23-12.05) for The Great Resilience Quest is out >> As our brave ...