Feature #18885
Updated by byroot (Jean Boussier) over 1 year ago
### Context Many optimizations in It is rather common to deploy Ruby with forking servers. A process first load the Ruby VM rely on lazily computed caches: Strings coderange, constant caches, method caches, etc etc. As such even without JIT, some operations need code and data of the application, and then forks a bit number of a warm up, workers to handle an incoming workload. The advantage is that each child has its own GVL and might be flushed if new constants are defined, new code its own GC, so they don't impact each others latency. The downside however is loaded, that in uses more memory than using threads or some objects are mutated. Additionally these lazily computed caches can cause fibers. That increased memory usage for applications relying is largely mitigated by Copy on Copy-on-Write memory. Whenever one of these caches is updated post fork, the entire Write, but it's far from perfect. Over time various memory page is invalidated. Precomputing these caches at the end of boot, even if based on heuristic, could improve Copy-on-Write performance. regions will be written into and unshared. The classic example is the objects generation, young objects must be promoted to the old generation before forking, otherwise they'll get invalidated on the next GC run. That's what https://github.com/ko1/nakayoshi_fork addresses. But there are other sources of CoW invalidation that could be addressed by MRI if it had a clear notification when it needs to be done. ### Proposal If applications had an API to notify the virtual machine that they're done loading code MRI could assume than any `fork` may be long lived and are about to start processing user input, it would give perform all the VM a good point in time to perform optimizations on the existing code and objects. e.g. could it can then, but It may be something like `RubyVM.prepare`, or `RubyVM.ready`. It's somewhat similar preferable to [Matz's static barrier idea from RubyConf 2020](https://youtu.be/JojpqfaPhjI?t=1908), except that it wouldn't disable any feature. have a dedicated API for that. e.g. - `Process.fork(long_lived: true)` - `Process.long_lived_fork` - `RubyVM.prepare_for_long_lived_fork` ### Potential optimizations `nakayoshi_fork` already does the following: - Do a major GC run to get rid of as many dangling objects as possible. - Promote all surviving objects to the highest generation - Compact the heap. But it would be much simpler to do this from inside the VM rather than do cryptic things such as `4.times { GC.start }` from the Ruby side. It's also not good to do this on every fork, once you fork the first long lived child, you shouldn't run it again. So decorating `fork` is not a good hook point. Also after discussing with @jhawthorn, @tenderlovemaking and @alanwu, we believe this would open the door to several other CoW optimizations: #### Precompute inline caches Even though we don't have hard data to prove it, we are convinced that a big source of CoW invalidation are inline caches. Most ISeq are never invoked during initialization, so child processed are forked with mostly cold caches. As a result the first time a method is executed in the child, many memory pages holding ISeq are invalidated as caches get updated. We think MRI could try to precompute these caches before forking children. Constant cache particularly should be resolvable statically see https://github.com/ruby/ruby/pull/6187. (somewhat related https://github.com/ruby/ruby/pull/6049). Method caches are harder to resolve statically, but we can probably apply some heuristics to at least reduce the cache misses. #### Copy on Write aware GC We could also keep some metadata about which memory pages are shared, or even introduce a "permanent" generation. [The Instagram engineering team introduced something like that in Python](https://instagram-engineering.com/copy-on-write-friendly-python-garbage-collection-ad6ed5233ddf) ([ticket](https://bugs.python.org/issue31558), [PR](https://github.com/python/cpython/pull/3705)). That makes the GC aware of which objects live on a shared page. With this information the GC can decide to no free dangling objects leaving on these pages, not to compact these pages, etc. #### Scan the coderange of all strings Strings have a lazily computed `coderange` attribute in their flags. So if a string is allocated at boot, but only used after fork, on first use its coderange will mayneed to may be computed and the string mutated. Using https://github.com/ruby/ruby/pull/6076, I noticed that 58% of the strings retained at the end of the boot sequence had an `UNKNOWN` coderange. So eagerly scanning the coderange of all strings could also improve Copy on Write performance. #### malloc_trim This hook will also be a good point to release unused pages to the system with `malloc_trim`.