Over the last year, I’ve been working on a new Ruby application server called Pitchfork. In most regards it’s extremely similar to the venerable Unicorn, from which it originates, but with one major...

Cache calls to Process.pid. Contribute to Shopify/pid_cache development by creating an account on GitHub.

An in-depth look at threads vs processes in Ruby web applications, and when you should use each.

Ruby 3.2 saw the introduction of object shapes which speedup instance variables access in most cases, but can be slower in some pathological cases.

It has been running Shopify's monolith for over 3 months now it is safe to assume major issues have been ironed out.

If the LHS is ASCII compatible and the RHS is 7BIT we can directly concat without being concerned about anything else. Benchmark: compare-ruby: ruby 3.2.0dev (2022-07-12T15:01:11Z master 71aec68566...

String#+@ is 2-3 times faster than String#dup because it can directly go through rb_str_dup instead of using the generic much slower rb_obj_dup. This fact led to the existence of the ugly Performan...

Fix: #8819 If the required capacity would fit in an embded string, returns one. This can reduce malloc churn for code that use string buffers. $ time ../miniruby-master -v -e '10_000_000.ti...

Previously the following read and wrote 1 byte out-of-bounds: $ valgrind ruby -e 'p /(\W+)[bx]\?/i.match? "aaaaaa aaaaaaaaa aaaa aaaaaaaa aaa aaaaxaaaaaaaaaaa aaaaa aaaaaaaaaaaa a ...

rb_str_concat does a lot of type checking we can easily bypass. | |compare-ruby|built-ruby| |:--------------|-----------:|---------:| |string_concat | 362.007k| 398.965k| | ...

https://kaigionrails.org/2023/talks/byroot/【発表概要】In this talk we'll go over two Rails bugs I fixed 10 years appart. For both I will detail how the bug was fo...

[Feature #18774] I created thread_sync.rb for easier handling of keyword arguments.

I recently got pulled on a memcached performance issue, we were seeing horribly slow memcached operations. After profiling it turns out that 99% of the time was spent in opentelemetry. This was par...

[Feature #19755] Before (in /tmp/test.rb): ```ruby Object.class_eval("p __FILE__") # => "(eval)" ``` After: ```ruby Object.class_eval("p __FI...

Ref: #1070 Ok, let's start the work to see what kind of roadblocks we'll run into. One idea I'm toying with right now, is that rather than delegate everything to redis-clien...

Using rb_str_buf_append instead of rb_enc_cr_str_buf_cat shows a 1.30x improvement on the micro benchmark. Skipping the method lookup in vm_objtostring shows and extra 1.10x, for a total of 1.40x.

BufferedIO is a bit inefficient for reading large responses because it use the classic buffer.slice! technique which cause a lot of unnecessary string copying. this is particularly visible on line ...

Set of GVL instrumentation tools. Contribute to Shopify/gvltools development by creating an account on GitHub.

Problem If you use Relation#extending or has_many extend:, the provided modules are extended onto the proxy instance. This means that every single instance has a different class and this is terribl...

It is common for network clients to use binary strings as buffer. However it is somehow sligthly slower than UTF-8 strings which makes no sense as in theory, less work is required. Benchmark: | ...

- Migrated to postgresql to be able to use `pg_sleep`, simulating a bit more IOs (extra 5ms per request) - Running with puma defaults: `bundle exec puma -e production --preload` - Benchmarking with...

Ruby gem to rescue from MySQL, PostgreSQL and Sqlite duplicate errors - Shopify/activerecord-rescue_from_duplicate