Using document _id

When I insert document to index Is it necessary to set valid _id for this document? I know that this document will not be updated in an index.

Just let Elasticsearch create a random one for you :slight_smile:

So document _id has valid value all the time. However when I checked document's _id in Kibana I didn't see any values of _id until I assign them directly

I'm not sure how that can be, every document gets assigned an _id automatically upon indexing into Elasticsearch.

@ybegus, did you verify that the time when the document was ingested, matches up with the Time Range filter in Kibana?

You should be able to see the document ID in the discover application in Kibana. It is usually the second field when a document is expanded.

Yes I did. So I saw my documents in Discover page in Kibana but _id fields were empty until I assigned values to _id in meta part of bulk function. As I understood it’s not necessary for user assign any values to _id if there is no plan to update documents, doesn’t it?

I agree with you. It’s easy to check. Actually I didn’t use _id field in meta-message.

Every document has to have an an _id set. Please show a screenshot where it is empty.

I have changed my Node.js code before I wrote about that and cleaned logs because I am in debugging process. I returned code back (removed _id from meta-data) to reproduce this behavior. However I see _id has value all the time like in a copy of one row from Discover page (see below). So I could not reproduce this behavior.

June 29th 2018, 19:22:14.335

ts:
June 29th 2018, 19:22:14.335
opID:
close
trqID:
10
pmID:
4294967306
vpID:
2
retCode:
0
read_bytes:
0
write_bytes:
17,600
fd:
1,455,553
user_name:
/developers/john
tpath:
/john-prj/vda/gendataTest/john_1601.txt
_id:
xMr-TGQBVw6i1FM-9AG8
_type:
log
_index:
icm-vda-20180629
_score:
-
Thank you for cooperation!

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.