Redisgraph: Need for GRAPH.BULK documentation

Created on 29 Oct 2019 · 3Comments · Source: RedisGraph/RedisGraph

Currently I am struggling with csv format of bulk loader utility. I have my own data format witch is binary and I don't wont to convert it to csv because the size of csv would be huge.
So I need to read documentation to write my own bulk loader

Documentation

Source

kkonevets

Most helpful comment

Any update on this? I see this is still undocumented which is disappointing.
Loading data in bulk is hugely important,

kevin-holbrook-epicor on 28 May 2020

👍3

All 3 comments

@kkonevets You are right, we don't currently have open documentation about the input format expected by the GRAPH.BULK endpoint that the bulk loader utility employs. It is not a secret, just ill-documented - the comments in the endpoint source code and utility code should be comprehensive in their description.

I'll augment our documentation with a full specification when possible, though it may be a week or so before time allows!

jeffreylovitz on 30 Oct 2019

😕2

Any update on this? I see this is still undocumented which is disappointing.
Loading data in bulk is hugely important,

kevin-holbrook-epicor on 28 May 2020

👍3

Also, it seems to be changing! I am using docker image redislabs/redisgraph:2.2.6

For example, https://github.com/RedisGraph/redisgraph-iam-demo is out of sync with it:

$ ./import-py3.sh
...
  File "redisgraph-bulk-loader/bulk_insert.py", line 270, in bulk_insert
    allocate_graph(label_descriptors.total_entities, relation_count)
  File "redisgraph-bulk-loader/bulk_insert.py", line 195, in allocate_graph
    result = redis_client.execute_command(*cmd)
...
redis.exceptions.ResponseError: Bulk insert format error, failed to parse bulk insert sections.