Matt Fleming

Pulsar async_api needs some kind of rate limiting to avoid long GC pauses

7

**Describe the bug** Using `async_api: true` in a Pulsar consumer scenario without rate limiting (e.g. the `rate=` param) causes the heap to fill up with millions of `CompletableFuture` objects because...

Inform the user that you can't use multiple Pulsar consumers with an exclusive subscription

2

**What is the ideal scenario for using this feature?** It would be helpful for users if they were told that using `threads= > 1` with an exclusive subscription type is...

tests: Add produce microbenchmark

5

## Backports Required - [x] none - not a bug fix - [ ] none - this is a backport - [ ] none - issue does not exist in...

Purge brokers no longer reported in metadata

10

Brokers that are not in the metadata should be purged from the internal client lists. This helps to avoid annoying "No route to host" and other connection failure messages. Fixes...

Fix for brokers with different Ids but same host:port

3

The Kafka protocol allows for brokers to have multiple host:port pairs for a given node Id, e.g. see UpdateMetadata request which contains a live_brokers list where each broker Id has...

It's helpful for consumers of the JSON output to know things like the mean before/after and the pvalue when assessing the severity of a change in performance.

`cargo bench` causes a panic

2

Running `cargo bench` with rustc 1.75.0 shows: ``` Benchmarking limbo/Execute prepared statement: 'SELECT * FROM users LIMIT 1': Warming up for 3.0000 sthread 'main' panicked at core/benches/benchmark.rs:48:25: internal error: entered...

Allow frontends to avoid jumping through hoops to create an Opt object, for example, when configuration is handled by a config file.

trie-free-latency: Demo showing the latency of freeing LPM_TRIE maps

2

The performance of freeing LPM_TRIE maps suffers heavily for maps with a large number of entries. Here's a little demo that measures the duration of trie_free() and allows the user...

process-util: Add support SCHED_EXT scheduling policy

5

Allow CPUSchedulingPolicy to be set to "ext". SCHED_EXT is a new scheduling policy in Linux v6.12 that allows processes to be scheduled using custom BPF schedulers instead of the default...