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
nwaku & postgres can work together when nwaku has the store protocol mounted.
In order to optimize the interaction with Postgres, we first need to have a testing tooling.
Details
Create a test project that can query a "store" node and analyze its performance.
Acceptance criteria
Have the initial test project with the essential stress test set. Of course, this project will grow over time.
Creation of the base test project
Test case: one single user publishes 1000 messages. All the messages should be stored properly. - [ ] Test case: 100 users publish 100 messages each, concurrently and each user sends 1 msg/sec. All the messages should be stored properly. Edited: that use case can't be implemented to validate Store performance because messages are dropped due to the Relay protocol (GossipSub nim-libp2p) not being able to support that high rate. For example, having 30 concurrent users publishing 300 messages concurrently will imply that a couple of messages can get lost. If we increase the number of users, then more messages can get lost.
The text was updated successfully, but these errors were encountered:
Background
nwaku
&postgres
can work together whennwaku
has thestore
protocol mounted.In order to optimize the interaction with
Postgres
, we first need to have a testing tooling.Details
Create a test project that can query a "store" node and analyze its performance.
Acceptance criteria
Have the initial test project with the essential stress test set. Of course, this project will grow over time.
- [ ] Test case: 100 users publish 100 messages each, concurrently and each user sends 1 msg/sec. All the messages should be stored properly.Edited: that use case can't be implemented to validate Store performance because messages are dropped due to the Relay protocol (GossipSub nim-libp2p) not being able to support that high rate. For example, having 30 concurrent users publishing 300 messages concurrently will imply that a couple of messages can get lost. If we increase the number of users, then more messages can get lost.The text was updated successfully, but these errors were encountered: