Hello,
I have an input config to monitor file output of Golden Gate (basically content of a Oracle database output as text file in real time). The file is continously write then switch to a new file in a chunk of 10'. The number of event is quite large, also the speed of writing (around 100mil events a day, ~600 events/sec at peak)
How do I know the log event is missing:
Because I counted the number of events in DB and counted the number of events in the log ingested. For example from 11:00:00 to 12:00:00, the ingested log have 1mil events, while the database have 2 mil rows.
The monitored config is as follow:
[monitor://outputfile/TEXT/RDB22/ISO/*/*/*/*] #The log file save in a partten of year/month/day/hour
disabled = false
sourcetype = mycustomsourcetype
index = mycustomindex
ignoreOlderThan = 20m
The sourcetype is as follow
[mycustomsourcetype]
SHOULD_LINEMERGE = false
LINE_BREAKER = (\r?\n)+I\|
INDEXED_EXTRACTIONS = CSV
FIELD_DELIMITER = |
HEADER_FIELD_LINE_NUMBER = 0
KV_MODE= none
TRUNCATE = 999999
TIME_PREFIX = \|(?=20\d{2}-\d{2}-\d{2}\s)
TIME_FORMAT = %Y-%m-%d %H:%M:%S
MAX_TIMESTAMP_LOOKAHEAD = 19
Any help with how to troubleshoot this problem would be very appriciate.
When you say you have checked the number of events in the log ingested, are you checking this for the number of events in Splunk or number of lines in the log file?
Assuming that an event is a single line in the log file, you could do something like this to get the number of events in the log file:
wc -l <pathToLog>This will provide the number of lines in the log, you should then to a check in Splunk:
| tstats count where index=yourIndex source=pathToLogPlease can you check these values and let us know what you get back?
🌟 Did this answer help you? If so, please consider:
Your feedback encourages the volunteers in this community to continue contributing