Profiling Particular person Queries in a Concurrent System

A great CPU profiler is value its weight in gold. Measuring efficiency in-situ often means utilizing a sampling profile. They supply plenty of data whereas having very low overhead. In a concurrent system, nonetheless, it’s laborious to make use of the ensuing knowledge to extract high-level insights. Samples don’t embrace context like question IDs and application-level statistics; they present you what code was run, however not why.

This weblog introduces trampoline histories, a method Rockset has developed to effectively connect application-level data (question IDs) to the samples of a CPU profile. This lets us use profiles to grasp the efficiency of particular person queries, even when a number of queries are executing concurrently throughout the identical set of employee threads.

Primer on Rockset

Rockset is a cloud-native search and analytics database. SQL queries from a buyer are executed in a distributed vogue throughout a set of servers within the cloud. We use inverted indexes, approximate vector indexes, and columnar layouts to effectively execute queries, whereas additionally processing streaming updates. The vast majority of Rockset’s performance-critical code is C++.

Most Rockset clients have their very own devoted compute assets known as digital cases. Inside that devoted set of compute assets, nonetheless, a number of queries can execute on the similar time. Queries are executed in a distributed vogue throughout the entire nodes, so because of this a number of queries are lively on the similar time in the identical course of. This concurrent question execution poses a problem when attempting to measure efficiency.

Concurrent question processing improves utilization by permitting computation, I/O, and communication to be overlapped. This overlapping is very vital for top QPS workloads and quick queries, which have extra coordination relative to their elementary work. Concurrent execution can be vital for decreasing head-of-line blocking and latency outliers; it prevents an occasional heavy question from blocking completion of the queries that observe it.

We handle concurrency by breaking work into micro-tasks which can be run by a hard and fast set of thread swimming pools. This considerably reduces the necessity for locks, as a result of we are able to handle synchronization through job dependencies, and it additionally minimizes context switching overheads. Sadly, this micro-task structure makes it tough to profile particular person queries. Callchain samples (stack backtraces) may need come from any lively question, so the ensuing profile exhibits solely the sum of the CPU work.

Profiles that mix the entire lively queries are higher than nothing, however plenty of guide experience is required to interpret the noisy outcomes. Trampoline histories allow us to assign a lot of the CPU work in our execution engine to particular person question IDs, each for steady profiles and on-demand profiles. This can be a very highly effective software when tuning queries or debugging anomalies.

DynamicLabel

The API we’ve constructed for including application-level metadata to the CPU samples is known as DynamicLabel. Its public interface may be very easy:

class DynamicLabel {
  public:
    DynamicLabel(std::string key, std::string worth);
    ~DynamicLabel();

    template <typename Func>
    std::invoke_result_t<Func> apply(Func&& func) const;
};

DynamicLabel::apply invokes func. Profile samples taken throughout that invocation could have the label connected.

Every question wants just one DynamicLabel. At any time when a micro-task from the question is run it’s invoked through DynamicLabel::apply.

Some of the vital properties of sampling profilers is that their overhead is proportional to their sampling charge; that is what lets their overhead be made arbitrarily small. In distinction, DynamicLabel::apply should do some work for each job whatever the sampling charge. In some circumstances our micro-tasks will be fairly micro, so it is necessary that apply has very low overhead.

apply‘s efficiency is the first design constraint. DynamicLabel‘s different operations (development, destruction, and label lookup throughout sampling) occur orders of magnitude much less regularly.

Let’s work by some methods we’d attempt to implement the DynamicLabel performance. We’ll consider and refine them with the objective of creating apply as quick as doable. If you wish to skip the journey and leap straight to the vacation spot, go to the “Trampoline Histories” part.

Implementation Concepts

Thought #1: Resolve dynamic labels at pattern assortment time

The obvious technique to affiliate software metadata with a pattern is to place it there from the start. The profiler would search for dynamic labels on the similar time that it’s capturing the stack backtrace, bundling a duplicate of them with the callchain.

Rockset’s profiling makes use of Linux’s perf_event, the subsystem that powers the perf command line software. perf_event has many benefits over signal-based profilers (reminiscent of gperftools). It has decrease bias, decrease skew, decrease overhead, entry to {hardware} efficiency counters, visibility into each userspace and kernel callchains, and the flexibility to measure interference from different processes. These benefits come from its structure, by which system-wide profile samples are taken by the kernel and asynchronously handed to userspace by a lock-free ring buffer.

Though perf_event has plenty of benefits, we are able to’t use it for concept #1 as a result of it may’t learn arbitrary userspace knowledge at sampling time. eBPF profilers have an analogous limitation.

Thought #2: Document a perf pattern when the metadata adjustments

If it’s not doable to tug dynamic labels from userspace to the kernel at sampling time, then what about push? We might add an occasion to the profile each time that the thread→label mapping adjustments, then post-process the profiles to match up the labels.

A method to do that can be to make use of perf uprobes. Userspace probes can document perform invocations, together with perform arguments. Sadly, uprobes are too sluggish to make use of on this vogue for us. Thread pool overhead for us is about 110 nanoseconds per job. Even a single crossing from the userspace into the kernel (uprobe or syscall) would multiply this overhead.

Avoiding syscalls throughout DynamicLabel::apply additionally prevents an eBPF resolution, the place we replace an eBPF map in apply after which modify an eBPF profiler like BCC to fetch the labels when sampling.

edit: eBPF can be utilized to tug from userspace when amassing a pattern, studying fsbase after which utilizing bpfprobelearnperson() to stroll a userspace knowledge construction that’s connected to a threadnative. When you’ve got BPF permissions enabled in your manufacturing atmosphere and are utilizing a BPF-based profiler then this different is usually a good one. The engineering and deployment points are extra advanced however the outcome doesn’t require in-process profile processing. Because of Jason Rahman for pointing this out.

Thought #3: Merge profiles with a userspace label historical past

If it is too costly to document adjustments to the thread→label mapping within the kernel, what if we do it within the userspace? We might document a historical past of calls to DynamicLabel::apply, then be a part of it to the profile samples throughout post-processing. perf_event samples can embrace timestamps and Linux’s CLOCK_MONOTONIC clock has sufficient precision to seem strictly monotonic (a minimum of on the x86_64 or arm64 cases we’d use), so the be a part of can be precise. A name to clock_gettime utilizing the VDSO mechanism is loads quicker than a kernel transition, so the overhead can be a lot decrease than that for concept #2.

The problem with this method is the information footprint. DynamicLabel histories can be a number of orders of magnitude bigger than the profiles themselves, even after making use of some easy compression. Profiling is enabled constantly on all of our servers at a low sampling charge, so attempting to persist a historical past of each micro-task invocation would rapidly overload our monitoring infrastructure.

Thought #4: In-memory historical past merging

The sooner we be a part of samples and label histories, the much less historical past we have to retailer. If we might be a part of the samples and the historical past in near-realtime (maybe each second) then we wouldn’t want to jot down the histories to disk in any respect.

The commonest means to make use of Linux’s perf_event subsystem is through the perf command line software, however the entire deep kernel magic is out there to any course of through the perf_event_open syscall. There are plenty of configuration choices (perf_event_open(2) is the longest manpage of any system name), however when you get it arrange you’ll be able to learn profile samples from a lock-free ring buffer as quickly as they’re gathered by the kernel.

To keep away from rivalry, we might preserve the historical past as a set of thread-local queues that document the timestamp of each DynamicLabel::apply entry and exit. For every pattern we might search the corresponding historical past utilizing the pattern’s timestamp.

This method has possible efficiency, however can we do higher?

Thought #5: Use the callchains to optimize the historical past of calls to `apply`

We are able to use the truth that apply exhibits up within the recorded callchains to scale back the historical past dimension. If we block inlining in order that we are able to discover DynamicLabel::apply within the name stacks, then we are able to use the backtrace to detect exit. Because of this apply solely wants to jot down the entry information, which document the time that an affiliation was created. Halving the variety of information halves the CPU and knowledge footprint (of the a part of the work that isn’t sampled).

This technique is the very best one but, however we are able to do even higher! The historical past entry information a variety of time for which apply was certain to a selected label, so we solely must make a document when the binding adjustments, relatively than per-invocation. This optimization will be very efficient if we have now a number of variations of apply to search for within the name stack. This leads us to trampoline histories, the design that we have now applied and deployed.

Trampoline Histories

If the stack has sufficient data to search out the appropriate DynamicLabel , then the one factor that apply must do is go away a body on the stack. Since there are a number of lively labels, we’ll want a number of addresses.

A perform that instantly invokes one other perform is a trampoline. In C++ it’d appear to be this:

__attribute__((__noinline__))
void trampoline(std::move_only_function<void()> func) {
    func();
    asm risky (""); // stop tailcall optimization
}

Word that we have to stop compiler optimizations that will trigger the perform to not be current within the stack, particularly inlining and tailcall elimination.

The trampoline compiles to solely 5 directions, 2 to arrange the body pointer, 1 to invoke func(), and a pair of to scrub up and return. Together with padding that is 32 bytes of code.

C++ templates allow us to simply generate an entire household of trampolines, every of which has a novel deal with.

utilizing Trampoline = __attribute__((__noinline__)) void (*)(
        std::move_only_function<void()>);

constexpr size_t kNumTrampolines = ...;

template <size_t N>
__attribute__((__noinline__))
void trampoline(std::move_only_function<void()> func) {
    func();
    asm risky (""); // stop tailcall optimization
}

template <size_t... Is>
constexpr std::array<Trampoline, sizeof...(Is)> makeTrampolines(
        std::index_sequence<Is...>) {
    return {&trampoline<Is>...};
}

Trampoline getTrampoline(unsigned idx) {
    static constexpr auto kTrampolines =
            makeTrampolines(std::make_index_sequence<kNumTrampolines>{});
    return kTrampolines.at(idx);
}

We’ve now bought the entire low-level items we have to implement DynamicLabel:

• DynamicLabel development → discover a trampoline that isn’t at present in use, append the label and present timestamp to that trampoline’s historical past
• DynamicLabel::apply → invoke the code utilizing the trampoline
• DynamicLabel destruction → return the trampoline to a pool of unused trampolines
• Stack body symbolization → if the trampoline’s deal with is present in a callchain, lookup the label within the trampoline’s historical past

Efficiency Affect

Our objective is to make DynamicLabel::apply quick, in order that we are able to use it to wrap even small items of labor. We measured it by extending our current dynamic thread pool microbenchmark, including a layer of indirection through apply.

{
    DynamicThreadPool executor({.maxThreads = 1});
    for (size_t i = 0; i < kNumTasks; ++i) {
        executor.add([&]() {
            label.apply([&] { ++rely; }); });
    }
    // ~DynamicThreadPool waits for all duties
}
EXPECT_EQ(kNumTasks, rely);

Maybe surprisingly, this benchmark exhibits zero efficiency impression from the additional stage of indirection, when measured utilizing both wall clock time or cycle counts. How can this be?

It seems we’re benefiting from a few years of analysis into department prediction for oblique jumps. The within of our trampoline appears like a digital technique name to the CPU. That is extraordinarily frequent, so processor distributors have put plenty of effort into optimizing it.

If we use perf to measure the variety of directions within the benchmark we observe that including label.apply causes about three dozen additional directions to be executed per loop. This might sluggish issues down if the CPU was front-end certain or if the vacation spot was unpredictable, however on this case we’re reminiscence certain. There are many execution assets for the additional directions, so that they don’t truly enhance this system’s latency. Rockset is mostly reminiscence certain when executing queries; the zero-latency outcome holds in our manufacturing atmosphere as effectively.

A Few Implementation Particulars

There are some things we have executed to enhance the ergonomics of our profile ecosystem:

• The perf.knowledge format emitted by perf is optimized for CPU-efficient writing, not for simplicity or ease of use. Though Rockset’s perf_event_open-based profiler pulls knowledge from perf_event_open, we have now chosen to emit the identical protobuf-based pprof format utilized by gperftools. Importantly, the pprof format helps arbitrary labels on samples and the pprof visualizer already has the flexibility to filter on these tags, so it was simple so as to add and use the data from DynamicLabel.
• We subtract one from most callchain addresses earlier than symbolizing, as a result of the return deal with is definitely the primary instruction that will likely be run after returning. That is particularly vital when utilizing inline frames, since neighboring directions are sometimes not from the identical supply perform.
• We rewrite trampoline<i> to trampoline<0> in order that we have now the choice of ignoring the tags and rendering an everyday flame graph.
• When simplifying demangled constructor names, we use one thing like Foo::copy_construct and Foo::move_construct relatively than simplifying each to Foo::Foo. Differentiating constructor sorts makes it a lot simpler to seek for pointless copies. (In case you implement this be sure to can deal with demangled names with unbalanced < and >, reminiscent of std::enable_if<sizeof(Foo) > 4, void>::sort.)
• We compile with -fno-omit-frame-pointer and use body tips to construct our callchains, however some vital glibc capabilities like memcpy are written in meeting and don’t contact the stack in any respect. For these capabilities, the backtrace captured by perf_event_open‘s PERF_SAMPLE_CALLCHAIN mode omits the perform that calls the meeting perform. We discover it by utilizing PERF_SAMPLE_STACK_USER to document the highest 8 bytes of the stack, splicing it into the callchain when the leaf is in a type of capabilities. That is a lot much less overhead than attempting to seize your complete backtrace with PERF_SAMPLE_STACK_USER.

Conclusion

Dynamic labels let Rockset tag CPU profile samples with the question whose work was lively at that second. This potential lets us use profiles to get insights about particular person queries, regardless that Rockset makes use of concurrent question execution to enhance CPU utilization.

Trampoline histories are a means of encoding the lively work within the callchain, the place the prevailing profiling infrastructure can simply seize it. By making the DynamicLabel ↔ trampoline binding comparatively long-lived (milliseconds, relatively than microseconds), the overhead of including the labels is stored extraordinarily low. The method applies to any system that wishes to reinforce sampled callchains with software state.

Rockset is hiring engineers in its Boston, San Mateo, London and Madrid places of work. Apply to open engineering positions immediately.