Trying to figure out how to scan from RedisPool

[abcd-ca]

Hi, I have a client variable which is of type RedisPool. I'm trying to call scan so that I can delete all keys that start with foo ("foo*"). It looks like I could call scan if my client was RedisClient but not if it's RedisPool but I'm not sure. I'd appreciate someone pointing me in the right direction. I'm expecting a lot of concurrent requests so that's why I'm going for a RedisPool over RedisClient.

[aembke]

@abcd-ca I added a more comprehensive scanning example here:

fred.rs/examples/scan.rs

Lines 50 to 92 in 31f8f56

	/// Example showing how to print the memory usage of all keys in a cluster with a `RedisPool`.
	///
	/// When using a clustered deployment the keyspace will be spread across multiple nodes. However, the cursor in each
	/// SCAN command is used to iterate over keys within a single node. There are several ways to concurrently scan
	/// all keys on all nodes:
	///
	/// 1. Use `scan_cluster`.
	/// 2. Use `split_cluster` and `scan`.
	/// 3. Use `with_cluster_node` and `scan`.
	///
	/// The best option depends on several factors, but `scan_cluster` is often the easiest approach for most use
	/// cases.
	async fn pool_scan_cluster_memory_example(pool: RedisPool) -> Result<(), RedisError> {
	// the majority of the client traffic in this scenario comes from the MEMORY USAGE call on each key, so we'll use a
	// pool to round-robin these commands among multiple clients. a clustered client with `auto_pipeline: true` can scan
	// all nodes in the cluster concurrently, so we use a single client rather than a pool to issue the SCAN calls.
	let mut total_size = 0;
	// if the pattern contains a hash tag then callers can use `scan` instead of `scan_cluster`
	let mut scanner = pool.next().scan_cluster("*", Some(100), None);
	while let Some(mut page) = scanner.try_next().await? {
	if let Some(page) = page.take_results() {
	// pipeline the `MEMORY USAGE` calls
	let pipeline = pool.next().pipeline();
	for key in page.iter() {
	let _: () = pipeline.memory_usage(key, Some(0)).await?;
	}
	let sizes: Vec<Option<u64>> = pipeline.all().await?;
	assert_eq!(page.len(), sizes.len());
	for (idx, key) in page.into_iter().enumerate() {
	let size = sizes[idx].unwrap_or(0);
	println!("{}: {}", key.as_str_lossy(), size);
	total_size += size;
	}
	}
	let _ = page.next();
	}
	println!("Total size: {}", total_size);
	Ok(())
	}

Hopefully that helps, but if you have any more questions feel free to reach out.

[abcd-ca]

Thank you for the follow-up :)