I am using Splunk to parse a particular sets of logs since many years but recently i have started facing some issue. Very few of the events are getting merged instead of parsing as a separate event.
Consider below example
2020-05-04 16:45:47,122 [ INFO] [CMEPS_JMSMessengerInject-EnterpriseMessageListener-186] - s_proc_id=921844e5-8130-4f29-9418-5622d95dfeef s_comp_id=ARCHIVER s_seq_no=9 s_proc_dur=372 s_proc_outcome=success
2020-05-04 16:45:48,124 [ INFO] [CMEPS_JMSMessengerInject-EnterpriseMessageListener-186]
These two events should be segregated and should not be merged under any condition.
Could someone help me to provide correct props.conf for this sourcetype so that Splunk only starts event with this timestamp pattern ex(2020-05-04 16:45:47)
One should never onboard a sourcetype without specific props.conf settings. Letting Splunk about the data is asking for problems and actually slows Splunk down. Every sourcetype should specify these six attributes:
TIME_PREFIX TIME_FORMAT MAX_TIMESTAMP_LOOKAHEAD SHOULD_LINEMERGE LINE_BREAKER TRUNCATE
If the data is sent by a universal forwarder then you should also specify
EVENT_BREAKER_ENABLE = true in the forwarder's props.conf.
Thank you so much for explaining about onboarding new sourcetype..so in present scenario which configuration you recommend, i just want my events to start with timestamp only with no merging of other timestamp event. Do I need to use breakonlybefore ?
Appreciate your response.
You should not need
BREAK_ONLY_BEFORE. Try these settings.
TIME_PREFIX = ^ TIME_FORMAT = %Y-%m-%d %H:%M:%S MAX_TIMESTAMP_LOOKAHEAD = 19 SHOULD_LINEMERGE = false LINE_BREAKER = ([\r\n]+) TRUNCATE = 10000