feat: aes-135-define-a-network-where-we-only-write-to-one-node-and-read #185
Conversation
samika98
changed the title
| if store_in_redis { | ||
| let mut conn: tokio::sync::MutexGuard<'_, MultiplexedConnection> = conn.lock().await; | ||
| let stream_id_string = response.to_string(); | ||
| let _: () = conn.sadd("anchor_mids", stream_id_string).await.unwrap(); |
There was a problem hiding this comment.
continuing my comment from above, if the connection is dead, we could grab user_data.redis_cli() and make/store a new one
runner/src/simulate.rs
Outdated
| peer: &Peer, | ||
| stream_id: String, | ||
| ) -> Result<bool, anyhow::Error> { | ||
| let client = reqwest::Client::new(); |
There was a problem hiding this comment.
creating a new client is heavy (it has an internal threadpool). we should get one of out of the user_data object/a local static/or pass it in as a parameter.
See https://docs.rs/reqwest/latest/reqwest/struct.Client.html specifically The Client holds a connection pool internally, so it is advised that you create one and reuse it.
There was a problem hiding this comment.
I do not see how we can use the GooseUser outside a GooseTransaction. We can create a static client and use that. For now I can create it in simulate.rs
| @@ -207,20 +237,24 @@ async fn instantiate_small_model( | |||
| user: &mut GooseUser, | |||
| store_in_redis: bool, | |||
| conn: Arc<tokio::sync::Mutex<MultiplexedConnection>>, | |||
| only_once_per_network: bool, | |||
There was a problem hiding this comment.
only_once_per_network confused me at first, I was thinking it was once globally, not one writer... maybe we can call it one_sided_test or one_writer_per_network or something? once makes me think literally once and that doesn't seem to be the intention.
| @@ -30,6 +30,7 @@ serde_ipld_dagcbor = "0.6" | |||
| serde_ipld_dagjson = "0.2" | |||
| schemars.workspace = true | |||
| serde_json.workspace = true | |||
| once_cell = "1.19.0" | |||
There was a problem hiding this comment.
nice, just an fyi you can use a OnceLock from the std library now, but I prefer the API on once_cell tbh
samika98
force-pushed
the
feature/aes-84-validate-sync-and-nightly
branch
from
4cc9317 to
df3a648
Compare
samika98
merged commit ccad158
into
feature/aes-84-validate-correctness-in-keramik-ceramic-anchoring-benchmark
Check the parent for description : #184