blob: 56f75688d928ce5fe8661a0cb2d0ada8f3bdea7f [file] [log] [blame]
Go GC: Prioritizing low latency and simplicity
31 Aug 2015
Richard Hudson
rlh@golang.org
* The Setup
Go is building a garbage collector (GC) not only for 2015 but for 2025 and
beyond: A GC that supports today’s software development and scales along with
new software and hardware throughout the next decade. Such a future has no
place for stop-the-world GC pauses, which have been an impediment to broader
uses of safe and secure languages such as Go.
Go 1.5, the first glimpse of this future, achieves GC latencies well below the
10 millisecond goal we set a year ago. We presented some impressive numbers
in [[https://talks.golang.org/2015/go-gc.pdf][a talk at Gophercon]].
The latency improvements have generated a lot of attention;
Robin Verlangen’s blog post
[[https://medium.com/@robin.verlangen/billions-of-request-per-day-meet-go-1-5-362bfefa0911][_Billions_of_requests_per_day_meet_Go_1.5_]]
validates our direction with end to end results.
We also particularly enjoyed
[[https://twitter.com/inconshreveable/status/620650786662555648][Alan Shreve’s production server graphs]]
and his "Holy 85% reduction" comment.
Today 16 gigabytes of RAM costs $100 and CPUs come with many cores, each with
multiple hardware threads. In a decade this hardware will seem quaint but the
software being built in Go today will need to scale to meet expanding needs and
the next big thing. Given that hardware will provide the power to increase
throughput, Go’s garbage collector is being designed to favor low latency and
tuning via only a single knob. Go 1.5 is the first big step down this path and
these first steps will forever influence Go and the applications it best
supports. This blog post gives a high-level overview of what we have done for
the Go 1.5 collector.
* The Embellishment
To create a garbage collector for the next decade, we turned to an algorithm
from decades ago. Go's new garbage collector is a _concurrent_, _tri-color_,
_mark-sweep_ collector, an idea first proposed by
[[http://dl.acm.org/citation.cfm?id=359655][Dijkstra in 1978]].
This is a deliberate divergence from most "enterprise" grade garbage collectors
of today, and one that we believe is well suited to the properties of modern
hardware and the latency requirements of modern software.
In a tri-color collector, every object is either white, grey, or black and we
view the heap as a graph of connected objects. At the start of a GC cycle all
objects are white. The GC visits all _roots_, which are objects directly
accessible by the application such as globals and things on the stack, and
colors these grey. The GC then chooses a grey object, blackens it, and then
scans it for pointers to other objects. When this scan finds a pointer to a
white object, it turns that object grey. This process repeats until there are
no more grey objects. At this point, white objects are known to be unreachable
and can be reused.
This all happens concurrently with the application, known as the _mutator_,
changing pointers while the collector is running. Hence, the mutator must
maintain the invariant that no black object points to a white object, lest the
garbage collector lose track of an object installed in a part of the heap it
has already visited. Maintaining this invariant is the job of the
_write_barrier_, which is a small function run by the mutator whenever a
pointer in the heap is modified. Go’s write barrier colors the now-reachable
object grey if it is currently white, ensuring that the garbage collector will
eventually scan it for pointers.
Deciding when the job of finding all grey objects is done is subtle and can be
expensive and complicated if we want to avoid blocking the mutators. To keep
things simple Go 1.5 does as much work as it can concurrently and then briefly
stops the world to inspect all potential sources of grey objects. Finding the
sweet spot between the time needed for this final stop-the-world and the total
amount of work that this GC does is a major deliverable for Go 1.6.
Of course the devil is in the details. When do we start a GC cycle? What
metrics do we use to make that decision? How should the GC interact with the Go
scheduler? How do we pause a mutator thread long enough to scan its stack?
 How do we represent white, grey, and black so we can efficiently find and scan
grey objects? How do we know where the roots are? How do we know where in an
object pointers are located? How do we minimize memory fragmentation? How do we
deal with cache performance issues? How big should the heap be? And on and on,
some related to allocation, some to finding reachable objects, some related to
scheduling, but many related to performance. Low-level discussions of each of
these areas are beyond the scope of this blog post.
At a higher level, one approach to solving performance problems is to add GC
knobs, one for each performance issue. The programmer can then turn the knobs
in search of appropriate settings for their application. The downside is that
after a decade with one or two new knobs each year you end up with the GC Knobs
Turner Employment Act. Go is not going down that path. Instead we provide a
single knob, called GOGC. This value controls the total size of the heap
relative to the size of reachable objects. The default value of 100 means that
total heap size is now 100% bigger than (i.e., twice) the size of the reachable
objects after the last collection. 200 means total heap size is 200% bigger
than (i.e., three times) the size of the reachable objects. If you want to
lower the total time spent in GC, increase GOGC. If you want to trade more GC
time for less memory, lower GOGC.
More importantly as RAM doubles with the next generation of hardware, simply
doubling GOGC will halve the number of GC cycles. On the other hand since GOGC
is based on reachable object size, doubling the load by doubling the reachable
objects requires no retuning. The application just scales.
Furthermore, unencumbered by ongoing support for dozens of knobs, the runtime
team can focus on improving the runtime based on feedback from real customer
applications.
* The Punchline
Go 1.5’s GC ushers in a future where stop-the-world pauses are no longer a
barrier to moving to a safe and secure language. It is a future where
applications scale effortlessly along with hardware and as hardware becomes
more powerful the GC will not be an impediment to better, more scalable
software. It’s a good place to be for the next decade and beyond.
For more details about the 1.5 GC and how we eliminated latency issues see the
[[https://www.youtube.com/watch?v=aiv1JOfMjm0][Go GC: Latency Problem Solved presentation]]
or [[https://talks.golang.org/2015/go-gc.pdf][the slides]].