You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It seems a good starting point for neptune sagemaker users is focusing on the magics & documenting around them
Describe the solution you'd like
Our current ideas are:
Inclusion as an optional viz tab of the jupyter magic's panel when registered for the %%gremlin, %%sparkl, etc calls
Check if similar ease-of-use for inline code query flows (direct. pygraphistry?) and NeptuneML flows
Instructions for enabling a neptune sagemaker vpc instance to talk to a graphistry server (self-hosted aws / marketplace, or Graphistry Hub). Not sure what is needed, just know users currently cannot use Sagemaker notebooks, even if a Graphistry server's nb instance can talk to Neptune
Happy to help on both and open to different approaches, continue in the Slack, etc
default settings sagemaker notebooks work w/ Graphistry Hub fine: outside internet
... but fails with graph-app-kit + aws marketplace private aws graphistry instances: sagemaker is https, while gak/marketplace instances are http b/c LetsEncrypt (autotls) & AWS do not work together. Maybe there is an AWS TLS thing graphistry-on-aws should auto-launch, at least for formation nodes?
enterprise neptune users likely have their own VPC issues, so maybe a diagnostic is what's needed for notebook networking and graphistry networking. given the user has jupyter, maybe a friendly script they can paste in that has the VPC/subnets and runs some connectivity tests?
Is your feature request related to a problem? Please describe.
The Graphistry team is seeing more use among Neptune users, and as their prototyping often starts in sagemaker, we're thinking of how to make their flow more natural than our normal jupyter integration ( https://github.com/graphistry/pygraphistry , the graph-app-kit launcher (https://github.com/graphistry/graph-app-kit/blob/master/docs/neptune.md), and our recent bindings
g.neptune(**creds).gremlin('query').plot()
(https://github.com/graphistry/pygraphistry/blob/master/demos/demos_databases_apis/neptune/neptune_tutorial.ipynb)It seems a good starting point for neptune sagemaker users is focusing on the magics & documenting around them
Describe the solution you'd like
Our current ideas are:
Inclusion as an optional viz tab of the jupyter magic's panel when registered for the
%%gremlin
,%%sparkl
, etc callsCheck if similar ease-of-use for inline code query flows (direct. pygraphistry?) and NeptuneML flows
Instructions for enabling a neptune sagemaker vpc instance to talk to a graphistry server (self-hosted aws / marketplace, or Graphistry Hub). Not sure what is needed, just know users currently cannot use Sagemaker notebooks, even if a Graphistry server's nb instance can talk to Neptune
Happy to help on both and open to different approaches, continue in the Slack, etc
cc @beebs-systap @bechbd
The text was updated successfully, but these errors were encountered: