I have two lookup tables created by a search with outputlookup command ,as:
table_1.csv with fields _time, A,B
table_2.csv with fields _time, A,C
I can use [|inputlookup table_1 ] and call the csv file ok.
I need to somehow join the two tables to get _time, A,B,C
NOTE: the common field in A
One fly in the ointment is that field A is not exclusive, by that I mean that it appears more than once in both tables, but within 30 seconds in both tables. So the time I need it to join is within A appearing within a 30s span of each other.
I need to join to be as one event (one line on the outputed csv file)
Then output to csv for a report which is scheduled daily.
just to share something i was looking for.
i had two lookups A and B with one field .and i wanted to know if value in lookup A was in lookup B.
so i rename the field by the same name 'c' (in the both lookups). i applied this search and i found what i wanted : |inputlookup A |join c [inputlookup B].
If your lookup tables have time, then it's possible that they should be indexed events and not lookup tables. In which case it would be possible to use the transaction command to issue the time restrictions you are looking for. This is still technically possible but a bit more awkward.
Try something like this:
| intputlookup table_1 | sort - _time | transaction A maxspan=30s | lookup table_2 A OUTPUT C | table _time, A, B, C
This assumes that the correct
max_matches setting for the definition of table_2. However, the date effectivity of the lookup will be dependent on on proper settings for
min_offset_secs. See the docs for transforms.conf for more details.
Another approach could be something like this:
| intputlookup table_1 | inputlookup append=t table_2 | sort - _time | transaction A maxpause=30s | table _time, A, B, C