boinc/doc/sched.php

<?php
require_once("docutil.php");
page_head("Client scheduling policies");
echo "

This document describes three related parts of the BOINC core client:
<dl>
<dt><b>CPU scheduling policy</b>
<dd>
Of the results that are runnable, which ones to execute?
BOINC will generally execute NCPUS results at once,
where NCPUS is the minimum of the physical number of CPUs
(counting hyperthreading) and the user's 'max_cpus' general preference.

<dt><b>CPU scheduling enforcement</b>
<dd>
When to actually enforce the schedule
(i.e. by preempting and starting tasks)?
Sometimes it's preferable to delay the preemption of
an application until it checkpoints.

<dt><b>Work-fetch policy</b>
<dd>
When should the core client ask a project for more work,
which project should it ask,
and how much work should it ask for?
</dl>

<p>
The goals of these policies are (in descending priority):
<ol>
<li> Results should be completed and reported by their deadline
(because results reported after their deadline
may not have any value to the project and may not be granted credit).
<li> NCPUS processors should be kept busy.
<li> At any given point, a computer should have enough work
so that NCPUS processors will be busy for at least
min_queue days (min_queue is a user preference).
<li> Project resource shares should be honored over the long term.
<li> If a computer is attached to multiple projects,
execution should rotate among projects on a frequent basis
(as defined by the user's 'CPU scheduling period' preference).
<li>
Execution should not switch between projects much more frequently
than the scheduling period,
Otherwise, if the 'remove processes from memory' preference is set,
and some applications take a long time to resume from a checkpoint,
lot of CPU time will be wasted.
</ol>

<p>
In previous versions of BOINC,
the core client attempted to maintain at least one result
for each attached project,
and would do weighted round-robin CPU scheduling among all projects.
In some scenarios (any combination of slow computer,
lots of projects, and tight deadlines) a computer could
miss the deadlines of all its results.
The new policies solve this problem as follows:
<ul>
<li>
Work fetch is limited to ensure that deadlines can be met.
A computer attached to 10 projects might
have work for only a few (perhaps only one) at a given time.
<li>
If deadlines are threatened,
the CPU scheduling policy optimizes the likelihood of meeting deadlines,
at the expense of variety.
</ul>

<h2>Concepts and terms</h2>

<h3>Wall CPU time</h3>
<p>
<b>Wall CPU time</b> is the amount of wall-clock time
a process has been runnable at the OS level.
The actual CPU time may be less than this,
e.g. if the process does a lot of paging,
or if other (non-BOINC) processing jobs run at the same time.
<p>
BOINC uses wall CPU time as the measure of CPU resource usage.
Wall CPU time is more fair than actual CPU time in the case of paging apps.
In addition, the measurement of actual CPU time depends on apps to
report it correctly, and they may not do this.

<h3>Normalized CPU time</h3>
<p>
The <b>normalized CPU time</b> of a result is an estimate
of the wall time it will take to complete, taking into account
<ul>
<li> the fraction of time BOINC runs ('on-fraction')
<li> the fraction of time computation is enabled ('active-fraction')
<li> CPU efficiency (the ratio of actual CPU to wall CPU)
</ul>
but not taking into account the project's resource share.

<h3>Project-normalized CPU time</h3>
<p>
The <b>project-normalized CPU time</b> of a result is an estimate
of the wall time it will take to complete, taking into account
the above factors plus the project's resource share
relative to other potentially runnable projects.
<p>
The 'work_req' element of a scheduler RPC request
is in units of project-normalized CPU time.
In deciding how much work to send,
the scheduler must take into account
the project's resource share fraction,
and the host's on-fraction and active-fraction.

<p>
For example, suppose a host has 1 GFLOP/sec CPUs,
the project's resource share fraction is 0.5,
the host's on-fraction is 0.8
and the host's active-fraction is 0.9.
Then the expected processing rate per CPU is

<pre>
(1 GFLOP/sec)*0.5*0.8*0.9 = 0.36 GFLOP/sec
</pre>

If the host requests 1000 project-normalized CPU seconds of work,
the scheduler should send it at least 360 GFLOPs of work.

<h3>Result states</h3>
R is <b>runnable</b> if
<ul>
<li> Neither R nor R.project is suspended, and
<li> R's input files have been downloaded, and
<li> R hasn't finished computing
</ul>

R is <b>nearly runnable</b> if
<ul>
<li> Neither R nor R.project is suspended, and
<li> None of R's input files is in a 'download deferred' state.
<li> R hasn't finished computing
</ul>

<h3>Project states</h3>
P is <b>runnable</b> if
<ul>
<li> P has at least one runnable result
(this implies that P is not suspended).
</ul>

P is <b>downloading</b> if
<ul>
<li> P is not suspended, and
<li> P has at least one result whose files are being downloaded
and none of the downloads is deferred.
</ul>

P is <b>fetchable</b>
(i.e. the work-fetch policy allows work to be fetched from it) if
<ul>
<li> P is not suspended, and
<li> P is not deferred (i.e. its minimum RPC time is in the past), and
<li> P's no-new-work flag is not set, and
<li> P is not overworked (see definition below), and
<li> a fetch of P's master file is not pending
</ul>

P is <b>latency-limited</b> if
<ul>
<li> The client's last scheduler RPC to P returned
a 'no work because of deadlines' flag, and
<li> the RPC reply's delay request has not yet elapsed.
</ul>
This means that P has work available,
but didn't send any because the work's deadlines couldn't be met
given the existing work queue.
<p>
P is <b>potentially runnable</b> if
<ul>
<li> P is either runnable, downloading, fetchable, overworked,
or latency-limited.
</ul>
This means that, to the best of the client's knowledge,
it could do work for P if it wanted to.

<h3>Debt</h3>
Intuitively, a project's 'debt' is how much work is owed to it,
relative to other projects.
BOINC uses two types of debt;
each is defined for a set S of projects.
In each case, the debt is recalculated periodically as follows:
<ul>
<li> A = the wall CPU time used by projects in S during this period
<li> R = sum of resource shares of projects in S
<li> For each project P in S:
   <ul>
   <li> F = P.resource_share / R (i.e., P's fractional resource share)
   <li> W = A*F (i.e., how much wall CPU time P should have gotten)
   <li> P.debt += W - P.wall_cpu_time (i.e. what P should have gotten
           minus what it got).
    </ul>
<li> P.debt is normalized so that the mean or minimum is zero.
</ul>


<b>Short-term debt</b> is used by the CPU scheduler.
It is adjusted over the set of runnable projects.
It is normalized so that minimum short-term debt is zero,
and maximum short-term debt is no greater than 86,400 (i.e. one day).

<p>
<b>Long-term debt</b> is used by the work-fetch policy.
It is defined for all projects,
and adjusted over the set of potentially runnable projects.
It is normalized so that average long-term debt,
over all project, is zero.

<h2>Round-robin simulation</h2>
<p>
The CPU scheduling and work fetch policies use the results
of a simulation of weighted round-robin scheduling
applied to the set of nearly runnable results.
The simulation takes into account on-fraction and active-fraction.
It produces the following outputs:
<ul>
<li> deadline_missed(R): whether result R misses its deadline.
<li> deadlines_missed(P):
the number of results R of P for which deadline_missed(R).
<li> total_shortfall:
the additional normalized CPU time needed to keep all CPUs busy
for the next min_queue seconds.
<li> shortfall(P):
the additional normalized CPU time needed for project P
to keep it from running out of work in the next min_queue seconds.
</ul>
<p>
In the example below, projects A and B have resource shares
2 and 1 respectively.
A has results A1 and A2, and B has result B1.
The computer has two CPUs.
From time 0 to 4 all three results run with equal weighting.
At time 4 result A2 finishes.
From time 4 to 8, project A gets only a 0.5 share
because it has only one result.
At time 8, result A1 finishes.
<p>
In this case, shortfall(A) is 4, shortfall(B) is 0, and total_shortfall is 2.

<br>
<img src=rr_sim.png>
<br>
<h2>CPU scheduling policy</h2>

<p>
The CPU scheduler uses an earliest-deadline-first (EDF) policy
for results that are in danger of missing their deadline,
and weighted round-robin among other projects if additional CPUs exist.
This allows the client to meet deadlines that would otherwise be missed,
while honoring resource shares over the long term.
The scheduling policy is:
<ol>
<li> Set the 'anticipated debt' of each project to its short-term debt
<li> Let P be the project with the earliest-deadline runnable result
among projects with deadlines_missed(P)>0.
Let R be P's earliest-deadline runnable result not scheduled yet.
Tiebreaker: least index in result array.
<li> If such an R exists, schedule R,
    decrement P's anticipated debt, and decrement deadlines_missed(P).
<li> If there are more CPUs, and projects with deadlines_missed(P)>0, go to 1.
<li> If all CPUs are scheduled, stop.
<li> If there is a result R that is currently running,
    and has been running for less than the CPU scheduling period,
    schedule R and go to 5.
<li> Find the project P with the greatest anticipated debt,
select one of P's runnable results
(picking one that is already running, if possible,
else the one received first from the project)
and schedule that result.
<li> Decrement P's anticipated debt by the 'expected payoff'
(the scheduling period divided by NCPUS).
<li> Go to 5.
</ol>

<p>
The CPU scheduler runs when a result is completed,
when the end of the user-specified scheduling period is reached,
when new results become runnable,
or when the user performs a UI interaction
(e.g. suspending or resuming a project or result).

<h2>CPU schedule enforcement</h2>
<p>
The CPU scheduler decides what results should run,
but it doesn't enforce this decision.
This enforcement is done by a separate
<b>scheduler enforcement function</b>,
which is called by the CPU scheduler at its conclusion.
Let X be the set of scheduled results that are not currently running,
let Y be the set of running results that are not scheduled,
and let T be the time the scheduler last ran.
The enforcement policy is as follows:
<ol>
<li> If deadline_missed(R) for some R in X,
then preempt a result in Y, and run R
(preempt the result with the least CPU wall time since checkpoint).
Repeat as needed.
<li> If there is a result R in Y that checkpointed more recently than T,
then preempt R and run a result in X.
</ol>



<h2>Work-fetch policy</h2>

<p>
A project P is <b>overworked</b> if
<ul>
<li> P.long_term_debt < -sched_period
</ul>
<p>
This condition occurs if P's results run in EDF mode
(and in extreme cases, when a project with large negative LTD is detached).
The work-fetch policy avoids getting work from overworked projects.
This prevents a situation where a project with short deadlines
gets more than its share of CPU time.

<p>
The work-fetch policy uses the functions
<pre>
frs(project P)
</pre>
<blockquote>
P's fractional resource share among fetchable projects.
</blockquote>

<p>
The work-fetch policy function is called every few minutes
(or as needed) by the scheduler RPC polling function.
It sets the variable <b>P.work_request_size</b> for each project P,
which is the number of seconds of work to request
if we do a scheduler RPC to P.
This is computed as follows:
<pre>
for each project P
    if P is suspended, deferred, overworked, or no-new-work
        P.work_request_size = 0
    else
        P.work_request_size = shortfall(P)

if total_shortfall > 0
    if P.work_request_size==0 for all P
        for each project P
            if P is suspended, deferred, overworked, or no-new-work
                continue
            P.work_request_size = 1

    if P.work_request_size==0 for all P
        for each project P
            if P is suspended, deferred, or no-new-work
                continue
            P.work_request_size = 1

    if P.work_request_size>0 for some P
        Normalize P.work_request_size so that they sum to total_shortfall
        and are proportional to P.resource_share
</pre>

<p>
For non-CPU-intensive projects,
P.work_request_size is set to 1 if P has no nearly-runnable result,
otherwise 0.
<p>
The scheduler RPC mechanism may select a project to contact
because of a user request, an outstanding trickle-up message,
or a result that is overdue for reporting.
If it does so, it will also request work from that project.
Otherwise, the RPC mechanism chooses the project P for which
<pre>
P.work_request_size>0 and
P.long_term_debt + shortfall(P) is greatest
</pre>
and requests work from that project.
Note: P.work_request_size is in units of normalized CPU time,
so the actual work request (which is in units of project-normalized CPU time)
is P.work_request_size
divided by P's resource share fraction relative to
potentially runnable projects.
<hr>
<a name=server_policy></a>
<h2>Scheduler work-send policy</h2>
<p>
NOTE: the following has not been implemented,
and is independent of the above policies.
<p>
The scheduler should avoid sending results whose
deadlines are likely to be missed,
or which are likely to cause existing results to miss their deadlines.
This will be accomplished as follows:
<ul>
<li>
Scheduler requests includes connection period,
list of queued result (with estimated time remaining and deadline)
and project resource fractions.
<li>
The scheduler won't send results whose deadlines are less than
now + min_queue.
<li>
The scheduler does an EDF simulation of the initial workload
to determine by how much each result misses its deadline.
For each result R being considered for sending,
the scheduler does an EDF simulation.
If R meets its deadline
and no result misses its deadline by more than it did previously, R is sent.

<li>
If the scheduler has work but doesn't send any because of deadline misses,
it returns a 'no work because of deadlines' flag.
If the last RPC to a project returned this flag,
it is marked as latency-limited and accumulates LTD.
</ul>
<hr>
<h2>Describing scenarios</h2>
<p>
We encourage the use of the following notation for
describing scheduling scenarios
(times are given in hours):
<p>
P(C, D, R)
<p>
This describes a project with
<ul>
<li> C = CPU time per task
<li> D = delay bound
<li> R = fractional resource share
</ul>

A scenario is described by a list of project,
plus the following optional parameters:
<ul>
<li> NCPUS: number of CPUS (default 1)
<li> min_queue
<li> leave_in_memory
<li> cpu_scheduling_period
</ul>

An example scenario description is:
<pre>
P1(1000, 2000, .5)
P2(1, 10, .5)
NCPUS=4
</pre>
<h2>Scenarios</h2>
<h3>Scenario 1</h3>
<p>
<pre>
P1(0.1, 1, .5)
P2(1, 24, .25)
P3(1, 24, .25)
NCPUS = 2
leave_in_memory = false
cpu_scheduling_period = 1
</pre>
Typically one CPU will process 6-minute tasks for P1,
and the other CPU will alternate between P2 and P3.
It's critical that the scheduler run each task of P2 and P3
for the full CPU scheduling period.
If we went strictly by debt, we'd end up switching between
them every 6 minutes,
and both P2 and P3 would have to resume from a checkpoint each time.
For some apps (e.g. Einstein@home) resuming from a checking
takes several minutes.
So we'd end up wasting most of the time on one CPU.
";
page_tail();
?>