Hello.
i'm posting to an HTTP endpoint and exporting results as JSON for auto extracting:
json_data = r.json()
while using 'addevent' and 'writeevents' to index results:
helper.addevent(json_data,sourcetype=eventSourceType)
helper.writeevents(index="main",host="localhost", source="sourcetype")
an error occurred:
Error: coercing to Unicode: need string or buffer, list found
only converting JSON to string resolves the issue, but it's preventing the results from being auto extracted.
is there a workaround other than convert JSON to string?
Thanks.
Usually when we got a big json, especially it's a list, we should iterate each element and call addevent() in a loop. However, if you want to compose a big event with all the elements, it's fine, Splunk will extract fields automatically only if it's a valid JSON.
Hey!
thanks for your comment.
my issue is not big json, but passing dict\list object to 'addevent' function.
even if i iterate it (see snip below). its still passing it as dict\list object. so error is still happening - signature="Error: coercing to Unicode: need string or buffer, list found.
passing it as string resolve it. the cons is the the results are not extracted automaticly
its all indexed as one line.
thanks
helper.addevent("",
sourcetype="sample_sourcetype")
for line in r.iter_lines():
if line:
print(json.loads(line))
helper.addevent(json.loads(line),
sourcetype=eventSourceType)
helper.writeevents(index="main", host="localhost", source="abuseipdb")
addevent() only accepts string as a param instead of JSON.
If it's a valid JSON, Splunk should be able to extract events automatically. If it's not, please try to add KV_MODE=json in your props.conf under the sourcetype stanza.