How to run the match match merge query in periodic iterate

[akshat_mittal]

Hello Team,

I need help in knowing how to run the MATCH MATCH MERGE query in periodic iterate.

Below is my query:

USING PERIODIC COMMIT 1000
LOAD CSV WITH HEADERS FROM "file:///granite2csv_path_ONT_CKT.csv" as row
MATCH (n:SPLITTER2_IN_PORT {SPLITTER2_NAME: row.PATH_ASIDE_EQUIPMENT})
MATCH (m:SPLITTER2_OUT_PORT {SP2_OUT_PORT_NAME: row.PATH_ASIDE_PORT})',
MERGE (n)-[r:SP2_INPORT_2_SP2_OUTPORT]->(m)
return r

Please help!

Regards
Akshat

[elaine_rosenber]

Hello Akshat,

Remove the comma after the second match. Returning the relationship is not very useful, but you can certainly do it,

Elaine

[akshat_mittal]

Hello Elaine,

Apologies for my mistake. , I forgot to remove that comma but in actual that comma is not present in my query. Hence I need to run the below query in periodic iterate.

Below query is taking too much time ( running from last 6 hours or elapsedTimeMillis = 20084612 ).

Please suggest better way to make this execution fast.

CSV contains 10 Lacs records.

USING PERIODIC COMMIT 1000
LOAD CSV WITH HEADERS FROM "file:///granite2csv_path_ONT_CKT.csv" as row
MATCH (n:SPLITTER2_IN_PORT {SPLITTER2_NAME: row.PATH_ASIDE_EQUIPMENT})
MATCH (m:SPLITTER2_OUT_PORT {SP2_OUT_PORT_NAME: row.PATH_ASIDE_PORT})
MERGE (n)-[r:SP2_INPORT_2_SP2_OUTPORT]->(m)
return r

Best Regards
Akshat

[akshat_mittal]

Hello Team ,

Please suggest on above query asked!

Thanks

Regards
Akshat

[andrew_bowman]

We don't recommend returning anything when doing a load (especially with PERIODIC COMMIT), try removing the RETURN.

Also, make sure you have an index on :SPLITTER2_IN_PORT(SPLITTER2_NAME) as well as on :SPLITTER2_OUT_PORT(SP2_OUT_PORT_NAME). You can double check using EXPLAIN to make sure you see two NodeIndexSeek operators and no NodeByLabelScan operators.