TSContScheduleOnPool¶
Synopsis¶
#include <ts/ts.h>
-
TSAction
TSContScheduleOnPool(TSCont contp, TSHRTime timeout, TSThreadPool tp)¶
Description¶
Schedules contp to run timeout milliseconds in the future, on a random thread that
belongs to tp. The timeout is an approximation, meaning it will be at least
timeout milliseconds but possibly more. Resolutions finer than roughly 5 milliseconds will
not be effective. Note that contp is required to have a mutex, which is provided to
TSContCreate().
The continuation is scheduled for a particular thread selected from a group of similar threads, as indicated by tp. If contp already has an thread affinity set, and the thread type of thread affinity is the same as tp, then contp will be scheduled on the thread specified by thread affinity.
Pool |
Properties |
|---|---|
|
Transaction processing threads. Continuations on these threads must not block. |
|
Background threads. Continuations can perform blocking operations. |
|
DNS request processing. May not exist depending on configuration. Not recommended. |
|
UDP processing. |
In practice, any choice except TS_THREAD_POOL_NET or TS_THREAD_POOL_TASK is strongly not
recommended. The TS_THREAD_POOL_NET threads are the same threads on which callback hooks are
called and continuations that use them have the same restrictions. TS_THREAD_POOL_TASK threads
are threads that exist to perform long or blocking actions, although sufficiently long operation can
impact system performance by blocking other continuations on the threads.
The return value can be used to cancel the scheduled event via TSActionCancel(). This is
effective until the continuation contp is being dispatched. However, if it is scheduled on
another thread this can be problematic to be correctly timed. The return value can be checked with
TSActionDone() to see if the continuation ran before the return, which is possible if
timeout is 0.
If contp has no thread affinity set, the thread it is now scheduled on will be set as its thread affinity thread.
Note that the TSContSchedule() family of API shall only be called from an ATS EThread. Calling it from raw non-EThreads can result in unpredictable behavior.
Note that some times if the TASK threads are not ready when you schedule a “contp” on the TS_THREAD_POOL_TASK then
the “contp” could end up being executed in a ET_NET thread instead, more likely this is not what you want.
To avoid this you can use the TS_LIFECYCLE_TASK_THREADS_READY_HOOK to make sure that the task threads
have been started when you schedule a “contp”. You can refer to TSLifecycleHookAdd() for more details.
Example Scenarios¶
Scenario 1 (no thread affinity info, different types of threads)¶
When thread affinity is not set, a plugin calls the API on thread “A” (which is an “ET_TASK” type), and wants to schedule on an “ET_NET” type thread provided in “tp”, the system would see there is no thread affinity information stored in “contp.”
In this situation, system sees there is no thread affinity information stored in “contp”. It then checks whether the type of thread “A” is the same as provided in “tp”, and sees that “A” is “ET_TASK”, but “tp” says “ET_NET”. So “contp” gets scheduled on the next available “ET_NET” thread provided by a round robin list, which we will call thread “B”. Since “contp” doesn’t have thread affinity information, thread “B” will be assigned as the affinity thread for it automatically.
The reason for doing this is most of the time people want to schedule the same things on the same type of thread, so logically it is better to default the first thread that it is scheduled on as the affinity thread.
Scenario 2 (no thread affinity info, same types of threads)¶
Slight variation of scenario 1, instead of scheduling on a “ET_NET” thread, the plugin wants to schedule on a “ET_TASK” thread (i.e. “tp” contains “ET_TASK” now), all other conditions stays the same.
This time since the type of the desired thread for scheduling and thread “A” are the same, the system schedules “contp” on thread “A”, and assigns thread “A” as the affinity thread for “contp”.
The reason behind this choice is that we are trying to keep things simple such that lock contention problems happens less. And for the most part, there is no point of scheduling the same thing on several different threads of the same type, because there is no parallelism between them (a thread will have to wait for the previous thread to finish, either because locking or the nature of the job it’s handling is serialized since its on the same continuation).
Scenario 3 (has thread affinity info, different types of threads)¶
Slight variation of scenario 1, thread affinity is set for continuation “contp” to thread “A”, all other conditions stays the same.
In this situation, the system sees that the “tp” has “ET_NET”, but the type of thread “A” is “ET_TASK”. So even though “contp” has an affinity thread, the system will not use that information since the type is different, instead it schedules “contp” on the next available “ET_NET” thread provided by a round robin list, which we will call thread “B”. The difference with scenario 1 is that since thread “A” is set to be the affinity thread for “contp” already, the system will NOT overwrite that information with thread “B”.
Most of the time, a continuation will be scheduled on one type of threads, and rarely gets scheduled on a different type. But when that happens, we want it to return to the thread it was previously on, so it won’t have any lock contention problems. And that’s also why “thread_affinity” is not a hashmap of thread types and thread pointers.
Scenario 4 (has thread affinity info, same types of threads)¶
Slight variation of scenario 3, the only difference is “tp” now says “ET_TASK”.
This is the easiest scenario since the type of thread “A” and “tp” are the same, so the system schedules “contp” on thread “A”. And, as discussed, there is really no reason why one may want to schedule the same continuation on two different threads of the same type.
Note
In scenario 3 & 4, it doesn’t matter which thread the plugin is calling the API from.