Redisgraph: redis-graph opencypher support and misc questions

Created on 21 Aug 2018 · 3Comments · Source: RedisGraph/RedisGraph

Hi, I just wanted to say that this project is awesome 💯 , after doing some testing I can see how amazingly fast it is ⚡️ 🙌

I have a couple of questions about redis-graph module:

Is it tested to work on cluster architecture?
I know about the clauses supported for opencypher, and I know SKIP is not supported yet, only LIMIT, but is there way to paginate without using SKIP?
Is it recommended to use redisgraph-py package?

Sidenote: I was hesitant to post in stackoverflow because there are only a few questions there, but if there is a better channel to ask these kind of things let me know, sorry for the inconvenient.

question

Source

p1nox

👍2

Most helpful comment

Glad you like the project!

Support for cluster architectures is on our roadmap, but is not implemented yet.
You're right that pagination is a bit tricky until we introduce SKIP clauses. As a temporary workaround, the easiest solution would be to include a unique identifier as a property in your nodes, then add a matching WHERE clause to make paginated queries:

CREATE (a:person {name: 'Jim', id: 1})
...
MATCH (a:person) WHERE a.id > 5 RETURN a ORDER BY a.id LIMIT 5

Soon we'll introduce logic to make this a bit less roundabout!

redisgraph-py is our most well-tested client at the moment, though we've recently introduced Java, Ruby, and Go clients as well - https://oss.redislabs.com/redisgraph/#client-libraries
I'd say use whatever is most natural to you and your use case, though! Queries can be passed to redis-cli on the command line, or through most Redis clients. The largest benefit the graph clients offer currently is more effectively being able to print and manipulate result sets.
I'm labeling this thread as question, but personally have no issue using GitHub as a forum for discussing the project. Thanks for your interest and your input!

jeffreylovitz on 21 Aug 2018

👍2

All 3 comments

Glad you like the project!

Support for cluster architectures is on our roadmap, but is not implemented yet.
You're right that pagination is a bit tricky until we introduce SKIP clauses. As a temporary workaround, the easiest solution would be to include a unique identifier as a property in your nodes, then add a matching WHERE clause to make paginated queries:

CREATE (a:person {name: 'Jim', id: 1})
...
MATCH (a:person) WHERE a.id > 5 RETURN a ORDER BY a.id LIMIT 5

Soon we'll introduce logic to make this a bit less roundabout!

redisgraph-py is our most well-tested client at the moment, though we've recently introduced Java, Ruby, and Go clients as well - https://oss.redislabs.com/redisgraph/#client-libraries
I'd say use whatever is most natural to you and your use case, though! Queries can be passed to redis-cli on the command line, or through most Redis clients. The largest benefit the graph clients offer currently is more effectively being able to print and manipulate result sets.
I'm labeling this thread as question, but personally have no issue using GitHub as a forum for discussing the project. Thanks for your interest and your input!

jeffreylovitz on 21 Aug 2018

👍2