This section describes the configuration of the condor_ startd to implement the desired policy for when remote jobs should start, be suspended, (possibly) resumed, vacate (with a checkpoint) or be killed (no checkpoint). This policy is the heart of Condor's balancing act between the needs and wishes of resource owners (machine owners) and resource users (people submitting their jobs to Condor). Please read this section carefully if you plan to change any of the settings described here, as a wrong setting can have a severe impact on either the owners of machines in your pool (they may ask to be removed from the pool entirely) or the users of your pool (they may stop using Condor).
Before we get into the details, there are a few things to note:
To define your policy, set expressions in the configuration file (see section 3.3 on Configuring Condor for an introduction to Condor's configuration files). The expressions are evaluated in the context of the machine's ClassAd and a job ClassAd. The expressions can therefore reference attributes from either ClassAd. Listed in this section are both the attributes that are included in the machine's ClassAd and the attributes that are included in a job ClassAd. The START expression is explained. It describes the conditions that must be met for a machine to start a job. The RANK expression is described. It allows the specification of the kinds of jobs a machine prefers to run. A final discussion details how the condor_ startd daemon works. Included are the machine states and activities, to give an idea of what is possible in policy decisions. Two example policy settings are presented.
The condor_ startd daemon represents the machine on which it is running to the Condor pool. The daemon publishes characteristics about the machine in the machine's ClassAd to aid matchmaking with resource requests. The values of these attributes may be listed by using the command: condor_ status -l hostname. On an SMP machine, the condor_ startd will break the machine up and advertise it as separate virtual machines, each with its own name and ClassAd. The attributes themselves and what they represent are described below:
X
for the VM with
name="vmX@full.hostname"For non-SMP machines with one virtual machine, the value will be 1.
In addition, there are a few attributes that are automatically inserted into the machine ClassAd whenever a resource is in the Claimed state:
There are a few attributes that are only inserted into the machine ClassAd if a job is currently executing. If the resource is claimed but no job are running, none of these attributes will be defined.
152.3
), as seen from condor_ q
on the submitting machine.
Finally, the single attribute, CurrentTime, is defined by the ClassAd environment.
CurrentTime - EnteredCurrentStatuswill equal the number of seconds that the job has been on hold.
Integer Code | Reason for Hold |
1 | The user put the job on hold with condor_ hold. |
2 | Globus middleware reported an error. HoldReasonSubCode is the GRAM error number. |
3 | The PERIODIC_HOLD expression evaluated to True. |
4 | The credentials for the job are invalid. |
5 | A job policy expression evaluated to Undefined. |
6 | The condor_ starter failed to start the executable. HoldReasonSubCode is the Unix error number. |
7 | The standard output file for the job could not be opened. HoldReasonSubCode is the Unix error number. |
8 | The standard input file for the job could not be opened. HoldReasonSubCode is the Unix error number. |
9 | The standard output stream for the job could not be opened. HoldReasonSubCode is the Unix error number. |
10 | The standard input stream for the job could not be opened. HoldReasonSubCode is the Unix error number. |
11 | An internal Condor protocol error was encountered when transferring files. |
12 | The condor_ starter failed to download input files. HoldReasonSubCode is the Unix error number. |
13 | The condor_ starter failed to upload output files. HoldReasonSubCode is the Unix error number. |
14 | The initial working directory of the job cannot be accessed. HoldReasonSubCode is the Unix error number. |
condor_q -hold
.
It can be used to determine if a job should be released or not.
Value | Status |
0 | Unexpanded (the job has never run) |
1 | Idle |
2 | Running |
3 | Removed |
4 | Completed |
5 | Held |
Value | Universe |
1 | standard |
4 | PVM |
5 | vanilla |
7 | scheduler |
8 | MPI |
9 | grid |
10 | java |
RemoteWallClockTime - CumulativeSuspensionTimeNote that this number does not get reset to zero when a job is forced to migrate from one machine to another.
The most important expression to the condor_ startd is the START expression. This expression describes the conditions that must be met for a machine to run a job. This expression can reference attributes in the machine's ClassAd (such as KeyboardIdle and LoadAvg) and attributes in a job ClassAd (such as Owner, Imagesize, and Cmd, the name of the executable the job will run). The value of the START expression plays a crucial role in determining the state and activity of a machine.
The Requirements expression is used for matching machines with jobs.
The condor_ startd defines the Requirements expression by logically anding the START expression and the IS_VALID_CHECKPOINT_PLATFORM expression.
In situations where a machine wants to make itself unavailable for further matches, the Requirements expression is set to FALSE. When the START expression locally evaluates to TRUE, the machine advertises the Requirements expression as TRUE and does not publish the START expression.
Normally, the expressions in the machine ClassAd are evaluated against certain request ClassAds in the condor_ negotiator to see if there is a match, or against whatever request ClassAd currently has claimed the machine. However, by locally evaluating an expression, the machine only evaluates the expression against its own ClassAd. If an expression cannot be locally evaluated (because it references other expressions that are only found in a request ad, such as Owner or Imagesize), the expression is (usually) undefined. See section 4.1 for specifics on how undefined terms are handled in ClassAd expression evaluation.
A note of caution is in order when modifying the START to reference job ClassAd attributes. The default IsOwner expression is a function of the START expression
START =?= FALSESee a detailed discussion of the IsOwner expression in section 3.5.8. However, the machine locally evaluates the IsOwner expression to determine if it is capable of running jobs for Condor. Any job ClassAd attributes appearing in the START expression, and hence in the IsOwner expression are undefined in this context, and may lead to unexpected behavior. Whenever the START expression is modified to reference job ClassAd attributes, the IsOwner expression should also be modified to reference only machine ClassAd attributes.
NOTE: If you have machines with lots of real memory and swap space such that the only scarce resource is CPU time, consider defining JOB_RENICE_INCREMENT so that Condor starts jobs on the machine with low priority. Then, further configure to set up the machines with:
START = True SUSPEND = False PREEMPT = False KILL = FalseIn this way, Condor jobs always run and can never be kicked off from activity on the machine. However, because they would run with ``nice priority'', interactive response on the machines will not suffer. You probably would not notice Condor was running the jobs, assuming you had enough free memory for the Condor jobs that there was little swapping.
A checkpoint is the platform-dependent information necessary to continue the execution of a standard universe job. Therefore, the machine (platform) upon which a job executed and produced a checkpoint limits the machines (platforms) which may use the checkpoint to continue job execution. This platform-dependent information is no longer the obvious combination of architecture and operating system, but may include subtle items such as the difference between the normal, bigmem, and hugemem kernels within the Linux operating system. This results in the incorporation of a separate expression to indicate the ability of a machine to resume and continue the execution of a job that has produced a checkpoint. The REQUIREMENTS expression is dependent on this information.
At a high level, IS_VALID_CHECKPOINT_PLATFORM is an expression which becomes true when a job's checkpoint platform matches the current checkpointing platform of the machine. Since this expression is anded with the START expression to produce the REQUIREMENTS expression, it must also behave correctly when evaluating in the context of jobs that are not standard universe.
In words, the current default policy for this expression:
Any non standard universe job may run on this machine. A standard universe job may run on machines with the new checkpointing identification system. A standard universe job may run if it has not yet produced a first checkpoint. If a standard universe job has produced a checkpoint, then make sure the checkpoint platforms between the job and the machine match.
The following is the default boolean expression for this policy. A JobUniverse value of 1 denotes the standard universe. This expression may be overridden in the Condor configuration files.
IS_VALID_CHECKPOINT_PLATFORM = ( ( (TARGET.JobUniverse == 1) == FALSE) || ( (MY.CheckpointPlatform =!= UNDEFINED) && ( (TARGET.LastCheckpointPlatform =?= MY.CheckpointPlatform) || (TARGET.NumCkpts == 0) ) ) )
IS_VALID_CHECKPOINT_PLATFORM is a separate policy expression because the complexity of IS_VALID_CHECKPOINT_PLATFORM can be very high. While this functionality is conceptually separate from the normal START policies usually constructed, it is also a part of the Requirements to allow the job to run.
A machine may be configured to prefer certain jobs over others using the RANK expression. It is an expression, like any other in a machine ClassAd. It can reference any attribute found in either the machine ClassAd or a request ad (normally, in fact, it references things in the request ad). The most common use of this expression is likely to configure a machine to prefer to run jobs from the owner of that machine, or by extension, a group of machines to prefer jobs from the owners of those machines.
For example, imagine there is a small research group with 4 machines called tenorsax, piano, bass, and drums. These machines are owned by the 4 users coltrane, tyner, garrison, and jones, respectively.
Assume that there is a large Condor pool in your department, but you spent a lot of money on really fast machines for your group. You want to implement a policy that gives priority on your machines to anyone in your group. To achieve this, set the RANK expression on your machines to reference the Owner attribute and prefer requests where that attribute matches one of the people in your group as in
RANK = Owner == "coltrane" || Owner == "tyner" \ || Owner == "garrison" || Owner == "jones"
The RANK expression is evaluated as a floating point number. However, like in C, boolean expressions evaluate to either 1 or 0 depending on if they are TRUE or FALSE. So, if this expression evaluated to 1 (because the remote job was owned by one of the preferred users), it would be a larger value than any other user (for whom the expression would evaluate to 0).
A more complex RANK expression has the same basic set up, where anyone from your group has priority on your machines. Its difference is that the machine owner has better priority on their own machine. To set this up for Jimmy Garrison, place the following entry in Jimmy Garrison's local configuration file bass.local:
RANK = (Owner == "coltrane") + (Owner == "tyner") \ + ((Owner == "garrison") * 10) + (Owner == "jones")NOTE: The parentheses in this expression are important, because ``+'' operator has higher default precedence than ``==''.
The use of ``+'' instead of ``| | '' allows us to distinguish which terms matched and which ones didn't. If anyone not in the John Coltrane quartet was running a job on the machine called bass, the RANK would evaluate numerically to 0, since none of the boolean terms evaluates to 1, and 0+0+0+0 still equals 0.
Suppose Elvin Jones submits a job. His job would match this machine (assuming the START was True for him at that time) and the RANK would numerically evaluate to 1. Therefore, Elvin would preempt the Condor job currently running. Assume that later Jimmy submits a job. The RANK evaluates to 10, since the boolean that matches Jimmy gets multiplied by 10. Jimmy would preempt Elvin, and Jimmy's job would run on Jimmy's machine.
The RANK expression is not required to reference the Owner of the jobs. Perhaps there is one machine with an enormous amount of memory, and others with not much at all. You can configure your large-memory machine to prefer to run jobs with larger memory requirements:
RANK = ImageSize
That's all there is to it. The bigger the job, the more this machine wants to run it. It is an altruistic preference, always servicing the largest of jobs, no matter who submitted them. A little less altruistic is John's RANK that prefers his jobs over those with the largest Imagesize:
RANK = (Owner == "coltrane" * 1000000000000) + ImagesizeThis RANK breaks if a job is submitted with an image size of more 1012 Kbytes. However, with that size, this RANK expression preferring that job would not be Condor's only problem!
A machine is assigned a state by Condor. The state depends on whether or not the machine is available to run Condor jobs, and if so, what point in the negotiations has been reached. The possible states are
Figure 3.3 shows the states and the possible transitions between the states.
Each transition is labeled with a letter. The cause of each transition is described below.
Within some machine states, activities of the machine are defined. The state has meaning regardless of activity. Differences between activities are significant. Therefore, a ``state/activity'' pair describes a machine. The following list describes all the possible state/activity pairs.
Figure 3.4 on
page gives the overall view of all
machine states and activities and shows the possible transitions
from one to another within the Condor system.
Each transition is labeled with a number on the diagram, and
transition numbers referred to in this manual will be bold.
Various expressions are used to determine when and if many of these state and activity transitions occur. Other transitions are initiated by parts of the Condor protocol (such as when the condor_ negotiator matches a machine with a schedd). The following section describes the conditions that lead to the various state and activity transitions.
This section traces through all possible state and activity transitions within a machine and describes the conditions under which each one occurs. Whenever a transition occurs, Condor records when the machine entered its new activity and/or new state. These times are often used to write expressions that determine when further transitions occurred. For example, enter the Killing activity if a machine has been in the Vacating activity longer than a specified amount of time.
When the startd is first spawned, the machine it represents enters the Owner state. The machine remains in the Owner state while the expression IsOwner is TRUE. If the IsOwner expression is FALSE, then the machine transitions to the Unclaimed state. The default value for the IsOwner expression is optimized for a shared resource
START =?= FALSESo, the machine will remain in the Owner state as long as the START expression locally evaluates to FALSE. Section 3.5.3 provides more detail on the START expression. If the START locally evaluates to TRUE or cannot be locally evaluated (it evaluates to UNDEFINED), transition 1 occurs and the machine enters the Unclaimed state. The IsOwner expression is locally evaluated by the machine, and should not reference job ClassAd attributes, which would be UNDEFINED.
For dedicated resources, the recommended value for the IsOwner expression is FALSE.
The Owner state represents a resource that is in use by its interactive owner (for example, if the keyboard is being used). The Unclaimed state represents a resource that is neither in use by its interactive user, nor the Condor system. From Condor's point of view, there is little difference between the Owner and Unclaimed states. In both cases, the resource is not currently in use by the Condor system. However, if a job matches the resource's START expression, the resource is available to run a job, regardless of if it is in the Owner or Unclaimed state. The only differences between the two states are how the resource shows up in condor_ status and other reporting tools, and the fact that Condor will not run benchmarking on a resource in the Owner state. As long as the IsOwner expression is TRUE, the machine is in the Owner State. When the IsOwner expression is FALSE, the machine goes into the Unclaimed State.
Here is an example that assumes that an IsOwner expression is not present in the configuration. If the START expression is
START = KeyboardIdle > 15 * $(MINUTE) && Owner == "coltrane"and if KeyboardIdle is 34 seconds, then the machine would remain in the Owner state. Owner is undefined, and
anything && FALSE
is FALSE.
If, however, the START expression is
START = KeyboardIdle > 15 * $(MINUTE) || Owner == "coltrane"and KeyboardIdle is 34 seconds, then the machine leaves the Owner state and becomes Unclaimed. This is because
FALSE || UNDEFINED
is UNDEFINED.
So, while this machine is not available to just anybody,
if user coltrane has jobs submitted, the machine is willing to run them.
Any other user's jobs have to wait
until KeyboardIdle exceeds 15 minutes.
However, since coltrane might claim this resource,
but has not yet, the machine goes to the Unclaimed state.
While in the Owner state, the startd polls the status of the machine every UPDATE_INTERVAL to see if anything has changed that would lead it to a different state. This minimizes the impact on the Owner while the Owner is using the machine. Frequently waking up, computing load averages, checking the access times on files, computing free swap space take time, and there is nothing time critical that the startd needs to be sure to notice as soon as it happens. If the START expression evaluates to TRUE and five minutes pass before the startd notices, that's a drop in the bucket of high-throughput computing.
The machine can only transition to the Unclaimed state from the Owner state. It does so when the IsOwner expression no longer evaluates to FALSE. By default, that happens when START no longer locally evaluates to FALSE.
Whenever the machine is not actively running a job, it will transition back to the Owner state if IsOwner evaluates to TRUE. Once a job is started, the value of IsOwner does not matter; the job either runs to completion or is preempted. Therefore, you must configure the preemption policy if you want to transition back to the Owner state from Claimed Busy.
If the IsOwner expression becomes TRUE, then the machine returns to the Owner state. If the IsOwner expression becomes FALSE, then the machine remains in the Unclaimed state. If the IsOwner expression is not present in the configuration files, then the default value for the IsOwner expression is
START =?= FALSEso that while in the Unclaimed state, if the START expression locally evaluates to FALSE, the machine returns to the Owner state by transition 2.
When in the Unclaimed state, the RunBenchmarks expression is relevant. If RunBenchmarks evaluates to TRUE while the machine is in the Unclaimed state, then the machine will transition from the Idle activity to the Benchmarking activity (transition 3) and perform benchmarks to determine MIPS and KFLOPS. When the benchmarks complete, the machine returns to the Idle activity (transition 4).
The startd automatically inserts an attribute, LastBenchmark, whenever it runs benchmarks, so commonly RunBenchmarks is defined in terms of this attribute, for example:
BenchmarkTimer = (CurrentTime - LastBenchmark) RunBenchmarks = $(BenchmarkTimer) >= (4 * $(HOUR))Here, a macro, BenchmarkTimer is defined to help write the expression. This macro holds the time since the last benchmark, so when this time exceeds 4 hours, we run the benchmarks again. The startd keeps a weighted average of these benchmarking results to try to get the most accurate numbers possible. This is why it is desirable for the startd to run them more than once in its lifetime.
NOTE: LastBenchmark is initialized to 0 before benchmarks have ever been run. To have the condor_ startd run benchmarks as soon as the machine is Unclaimed (if it has not done so already), include a term using LastBenchmark as in the example above.
NOTE: If RunBenchmarks is defined and set to something other than FALSE, the startd will automatically run one set of benchmarks when it first starts up. To disable benchmarks, both at startup and at any time thereafter, set RunBenchmarks to FALSE or comment it out of the configuration file.
From the Unclaimed state, the machine can go to four other possible states: Owner (transition 2), Backfill/Idle, Matched, or Claimed/Idle.
Once the condor_ negotiator matches an Unclaimed machine with a requester at a given schedd, the negotiator sends a command to both parties, notifying them of the match. If the schedd receives that notification and initiates the claiming procedure with the machine before the negotiator's message gets to the machine, the Match state is skipped, and the machine goes directly to the Claimed/Idle state (transition 5). However, normally the machine will enter the Matched state (transition 6), even if it is only for a brief period of time.
If the machine has been configured to perform backfill jobs (see section 3.13.9), while it is in Unclaimed/Idle it will evaluate the START_BACKFILL expression. Once START_BACKFILL evaluates to TRUE, the machine will enter the Backfill/Idle state (transition 7) to begin the process of running backfill jobs.
The Matched state is not very interesting to Condor. Noteworthy in this state is that the machine lies about its START expression while in this state and says that Requirements are false to prevent being matched again before it has been claimed. Also interesting is that the startd starts a timer to make sure it does not stay in the Matched state too long. The timer is set with the MATCH_TIMEOUT configuration file macro. It is specified in seconds and defaults to 120 (2 minutes). If the schedd that was matched with this machine does not claim it within this period of time, the machine gives up, and goes back into the Owner state via transition 8. It will probably leave the Owner state right away for the Unclaimed state again and wait for another match.
At any time while the machine is in the Matched state, if the START expression locally evaluates to FALSE, the machine enters the Owner state directly (transition 8).
If the schedd that was matched with the machine claims it before the MATCH_TIMEOUT expires, the machine goes into the Claimed/Idle state (transition 9).
The Claimed state is certainly the most complex state. It has the most possible activities and the most expressions that determine its next activities. In addition, the condor_ checkpoint and condor_ vacate commands affect the machine when it is in the Claimed state. In general, there are two sets of expressions that might take effect. They depend on the universe of the request: standard or vanilla. The standard universe expressions are the normal expressions. For example:
WANT_SUSPEND = True WANT_VACATE = $(ActivationTimer) > 10 * $(MINUTE) SUSPEND = $(KeyboardBusy) || $(CPUBusy) ...
The vanilla expressions have the string``_VANILLA'' appended to their names. For example:
WANT_SUSPEND_VANILLA = True WANT_VACATE_VANILLA = True SUSPEND_VANILLA = $(KeyboardBusy) || $(CPUBusy) ...
Without specific vanilla versions, the normal versions will be used for all jobs, including vanilla jobs. In this manual, the normal expressions are referenced. The difference exists for the the resource owner that might want the machine to behave differently for vanilla jobs, since they cannot checkpoint. For example, owners may want vanilla jobs to remain suspended for longer than standard jobs.
While Claimed, the POLLING_INTERVAL takes effect, and the startd polls the machine much more frequently to evaluate its state.
If the machine owner starts typing on the console again, it is best to notice this as soon as possible to be able to start doing whatever the machine owner wants at that point. For SMP machines, if any virtual machine is in the Claimed state, the startd polls the machine frequently. If already polling one virtual machine, it does not cost much to evaluate the state of all the virtual machines at the same time.
There are a variety of events that may cause the startd to try to get rid of or temporarily suspend a running job. Activity on the machine's console, load from other jobs, or shutdown of the startd via an administrative command are all possible sources of interference. Another one is the appearance of a higher priority claim to the machine by a different Condor user.
Depending on the configuration, the startd may respond quite
differently to activity on the machine, such as keyboard activity or
demand for the cpu from processes that are not managed by Condor. The
startd can be configured to completely ignore such activity or to
suspend the job or even to kill it. A standard configuration for a desktop
machine might be to go through
successive levels of getting the job out of the way.
The first and least costly to the job is suspending it.
This works for both standard and vanilla jobs.
If suspending the job for a short while does not satisfy the machine
owner (the owner is still using the machine after a specific period of
time), the startd moves on to vacating the job.
Vacating a standard universe job
involves performing a checkpoint so that the work already completed
is not lost. Vanilla jobs are sent a soft kill signal so that they
can gracefully shut down if necessary; the default is SIGTERM
.
If vacating does not satisfy the machine owner (usually because it is
taking too long and the owner wants their machine back now),
the final, most drastic stage is reached: killing.
Killing is a quick death to the job, using a hard-kill signal that cannot
be intercepted by the application. For vanilla jobs that do no special
signal handling, vacating and killing are equivalent.
The WANT_SUSPEND expression determines if the machine will evaluate the SUSPEND expression to consider entering the Suspended activity. The WANT_VACATE expression determines what happens when the machine enters the Preempting state. It will go to the Vacating activity or directly to Killing. If one or both of these expressions evaluates to FALSE, the machine will skip that stage of getting rid of the job and proceed directly to the more drastic stages.
When the machine first enters the Claimed state, it goes to the Idle activity. From there, it has two options. It can enter the Preempting state via transition 10 (if a condor_ vacate arrives, or if the START expression locally evaluates to FALSE), or it can enter the Busy activity (transition 11) if the schedd that has claimed the machine decides to activate the claim and start a job.
From Claimed/Busy, the machine can transition to three other state/activity pairs. The startd evaluates the WANT_SUSPEND expression to decide which other expressions to evaluate. If WANT_SUSPEND is TRUE, then the startd evaluates the SUSPEND expression. If WANT_SUSPEND is FALSE, then the startd will evaluate the PREEMPT expression and skip the Suspended activity entirely. By transition, the possible state/activity destinations from Claimed/Busy:
Another reason the machine would go from Claimed/Busy to Claimed/Retiring is if the condor_ negotiator matched the machine with a ``better'' match. This better match could either be from the machine's perspective using the startd RANK expression, or it could be from the negotiator's perspective due to a job with a higher user priority.
Another case resulting in a transition to Claimed/Retiring is when the startd is being shut down. The only exception is a ``fast'' shutdown, which bypasses retirement completely.
If a condor_ checkpoint command arrives, or the PeriodicCheckpoint expression evaluates to TRUE, there is no state change. The startd has no way of knowing when this process completes, so periodic checkpointing can not be another state. Periodic checkpointing remains in the Claimed/Busy state and appears as a running job.
From the Claimed/Suspended state, the following transitions may occur:
For the Claimed/Retiring state, the following transitions may occur:
The Preempting state is less complex than the Claimed state. There are two activities. Depending on the value of WANT_VACATE, a machine will be in the Vacating activity (if TRUE) or the Killing activity (if FALSE).
While in the Preempting state (regardless of activity) the machine advertises its Requirements expression as FALSE to signify that it is not available for further matches, either because it is about to transition to the Owner state, or because it has already been matched with one preempting match, and further preempting matches are disallowed until the machine has been claimed by the new match.
The main function of the Preempting state is to get rid of the starter associated with the resource. If the condor_ starter associated with a given claim exits while the machine is still in the Vacating activity, then the job successfully completed a graceful shutdown. For standard universe jobs, this means that a checkpoint was saved. For other jobs, this means the application was given an opportunity to do a graceful shutdown, by intercepting the soft kill signal.
If the machine is in the Vacating activity, it keeps evaluating the KILL expression. As soon as this expression evaluates to TRUE, the machine enters the Killing activity (transition 21).
When the starter exits, or if there was no starter running when the machine enters the Preempting state (transition 10), the other purpose of the Preempting state is completed: notifying the schedd that had claimed this machine that the claim is broken.
At this point, the machine enters either the Owner state by transition 22 (if the job was preempted because the machine owner came back) or the Claimed/Idle state by transition 23 (if the job was preempted because a better match was found).
If the machine enters the Killing activity, (because either WANT_VACATE was FALSE or the KILL expression evaluated to TRUE), it attempts to force the condor_ starter to immediately kill the underlying Condor job. Once the machine has begun to hard kill the Condor job, the condor_ startd starts a timer, the length of which is defined by the KILLING_TIMEOUT macro. This macro is defined in seconds and defaults to 30. If this timer expires and the machine is still in the Killing activity, something has gone seriously wrong with the condor_ starter and the startd tries to vacate the job immediately by sending SIGKILL to all of the condor_ starter's children, and then to the condor_ starter itself.
Once the condor_ starter has killed off all the processes associated with the job and exited, and once the schedd that had claimed the machine is notified that the claim is broken, the machine will leave the Preempting/Killing state. If the job was preempted because a better match was found, the machine will enter Claimed/Idle (transition 24). If the preemption was caused by the machine owner (the PREEMPT expression evaluated to TRUE, condor_ vacate was used, etc), the machine will enter the Owner state (transition 25).
The Backfill state is used whenever the machine is performing low
priority background tasks to keep itself busy.
For more information about backfill support in Condor, see
section 3.13.9 on page .
This state is only used if the machine has been configured to enable
backfill computation, if a specific backfill manager has been
installed and configured, and if the machine is otherwise idle (not
being used interactively or for regular Condor computations).
If the machine meets all these requirements, and the
START_BACKFILL expression evalutes to TRUE, the machine will
move from the Unclaimed/Idle state to Backfill/Idle (transition
7).
Once a machine is in Backfill/Idle, it will immediately attempt to spawn whatever backfill manager it has been configured to use (currently, only the BOINC client is supported as a backfill manager in Condor). Once the BOINC client is running, the machine will enter Backfill/Busy (transition 26) to indicate that it is now performing a backfill computation.
NOTE: On SMP machines, the condor_ startd will only spawn a single instance of the BOINC client, even if multiple virtual machines are available to run backfill jobs. Therefore, only the first machine to enter Backfill/Idle will cause a copy of the BOINC client to start running. If a given virtual machine on an SMP enters the Backfill state and a BOINC client is already running under this condor_ startd, the virtual machine will immediately enter Backfill/Busy without waiting to spawn another copy of the BOINC client.
If the BOINC client ever exits on its own (which normally wouldn't happen), the machine will go back to Backfill/Idle (transition 27) where it will immediately attempt to respawn the BOINC client (and return to Backfill/Busy via transition 26).
As the BOINC client is running a backfill computation, a number of events can occur that will drive the machine out of the Backfill state. The machine can get matched or claimed for a Condor job, interactive users can start using the machine again, the machine might be evicted with condor_ vacate, or the condor_ startd might be shutdown. All of these events cause the condor_ startd to kill the BOINC client and all its descendants, and enter the Backfill/Killing state (transition 28).
Once the BOINC client and all its children have exited the system, the machine will enter the Backfill/Idle state to indicate that the BOINC client is now gone (transition 29). As soon as it enters Backfill/Idle after the BOINC client exits, the machine will go into another state, depending on what caused the BOINC client to be killed in the first place.
If the EVICT_BACKFILL expression evaluates to TRUE while a machine is in Backfill/Busy, after the BOINC client is gone, the machine will go back into the Owner/Idle state (transition 30). The machine will also return to the Owner/Idle state after the BOINC client exits if condor_ vacate was used, or if the condor_ startd is being shutdown.
When a machine running backfill jobs is matched with a requester that wants to run a Condor job, the machine will either enter the Matched state, or go directly into Claimed/Idle. As with the case of a machine in Unclaimed/Idle (described above), the condor_ negotiator informs both the condor_ startd and the condor_ schedd of the match, and the exact state transitions at the machine depend on what order the various entities initiate communication with each other. If the condor_ schedd is notified of the match and sends a request to claim the condor_ startd before the condor_ negotiator has a chance to notify the condor_ startd, once the BOINC client exits, the machine will immediately enter Claimed/Idle (transition 31). Normally, the notification from the condor_ negotiator will reach the condor_ startd before the condor_ schedd attempts to claim it. In this case, once the BOINC client exits, the machine will enter Matched/Idle (transition 32).
Once the job finishes or if the retirement time expires, the machine enters the Preempting state.
This section describes the default configuration policy and then provides examples of extensions to these policies.
These settings are the default as shipped with Condor. They have been used for many years with no problems. The vanilla expressions are identical to the regular ones. (They are not listed here. If not defined, the standard expressions are used for vanilla jobs as well).
The following are macros to help write the expressions clearly.
## These macros are here to help write legible expressions: MINUTE = 60 HOUR = (60 * $(MINUTE)) StateTimer = (CurrentTime - EnteredCurrentState) ActivityTimer = (CurrentTime - EnteredCurrentActivity) ActivationTimer = (CurrentTime - JobStart) LastCkpt = (CurrentTime - LastPeriodicCheckpoint) NonCondorLoadAvg = (LoadAvg - CondorLoadAvg) BackgroundLoad = 0.3 HighLoad = 0.5 StartIdleTime = 15 * $(MINUTE) ContinueIdleTime = 5 * $(MINUTE) MaxSuspendTime = 10 * $(MINUTE) MaxVacateTime = 10 * $(MINUTE) KeyboardBusy = KeyboardIdle < $(MINUTE) ConsoleBusy = (ConsoleIdle < $(MINUTE)) CPUIdle = $(NonCondorLoadAvg) <= $(BackgroundLoad) CPUBusy = $(NonCondorLoadAvg) >= $(HighLoad) KeyboardNotBusy = ($(KeyboardBusy) == False) MachineBusy = ($(CPUBusy) || $(KeyboardBusy)
Macros are defined to want to suspend jobs (instead of killing them) in the case of jobs that use little memory, when the keyboard is not being used, and for vanilla universe and PVM universe jobs. We want to gracefully vacate jobs which have been running for more than 10 minutes or are vanilla universe or PVM universe jobs.
WANT_SUSPEND = ( $(SmallJob) || $(KeyboardNotBusy) \ || $(IsPVM) || $(IsVanilla) ) WANT_VACATE = ( $(ActivationTimer) > 10 * $(MINUTE) \ || $(IsPVM) || $(IsVanilla) )
Finally, definitions of the actual expressions. Start a job if the keyboard has been idle long enough and the load average is low enough OR the machine is currently running a Condor job. Note that Condor would only run one job at a time. It just may prefer to run a different job, as defined by the machine rank or user priorities.
START = ( (KeyboardIdle > $(StartIdleTime)) \ && ( $(CPUIdle) || \ (State != "Unclaimed" && State != "Owner")) )
Suspend a job if the keyboard has been touched. Alternatively, suspend if the CPU has been busy for more than two minutes and the job has been running for more than 90 seconds.
SUSPEND = ( $(KeyboardBusy) || \ ( (CpuBusyTime > 2 * $(MINUTE)) \ && $(ActivationTimer) > 90 ) )
Continue a suspended job if the CPU is idle, the Keyboard has been idle for long enough, and the job has been suspended more than 10 seconds.
CONTINUE = ( $(CPUIdle) && ($(ActivityTimer) > 10) \ && (KeyboardIdle > $(ContinueIdleTime)) )
There are two conditions that signal preemption. The first condition is if the job is suspended, but it has been suspended too long. The second condition is if suspension is not desired and the machine is busy.
PREEMPT = ( ((Activity == "Suspended") && \ ($(ActivityTimer) > $(MaxSuspendTime))) \ || (SUSPEND && (WANT_SUSPEND == False)) )
Do not give jobs any time to retire on their own when they are about to be preempted.
MaxJobRetirementTime = 0
Kill jobs that take too long leaving gracefully.
KILL = $(ActivityTimer) > $(MaxVacateTime)
Finally, specify periodic checkpointing. For jobs smaller than 60 Mbytes, do a periodic checkpoint every 6 hours. For larger jobs, only checkpoint every 12 hours.
PERIODIC_CHECKPOINT = ( (ImageSize < 60000) && \ ($(LastCkpt) > (6 * $(HOUR))) ) || \ ( $(LastCkpt) > (12 * $(HOUR)) )
At UW-Madison, we have a fast network. We simplify our expression considerably to
PERIODIC_CHECKPOINT = $(LastCkpt) > (3 * $(HOUR))
For reference, the entire set of policy settings are included once more without comments:
## These macros are here to help write legible expressions: MINUTE = 60 HOUR = (60 * $(MINUTE)) StateTimer = (CurrentTime - EnteredCurrentState) ActivityTimer = (CurrentTime - EnteredCurrentActivity) ActivationTimer = (CurrentTime - JobStart) LastCkpt = (CurrentTime - LastPeriodicCheckpoint) NonCondorLoadAvg = (LoadAvg - CondorLoadAvg) BackgroundLoad = 0.3 HighLoad = 0.5 StartIdleTime = 15 * $(MINUTE) ContinueIdleTime = 5 * $(MINUTE) MaxSuspendTime = 10 * $(MINUTE) MaxVacateTime = 10 * $(MINUTE) KeyboardBusy = KeyboardIdle < $(MINUTE) ConsoleBusy = (ConsoleIdle < $(MINUTE)) CPUIdle = $(NonCondorLoadAvg) <= $(BackgroundLoad) CPUBusy = $(NonCondorLoadAvg) >= $(HighLoad) KeyboardNotBusy = ($(KeyboardBusy) == False) MachineBusy = ($(CPUBusy) || $(KeyboardBusy) WANT_SUSPEND = ( $(SmallJob) || $(KeyboardNotBusy) \ || $(IsPVM) || $(IsVanilla) ) WANT_VACATE = ( $(ActivationTimer) > 10 * $(MINUTE) \ || $(IsPVM) || $(IsVanilla) ) START = ( (KeyboardIdle > $(StartIdleTime)) \ && ( $(CPUIdle) || \ (State != "Unclaimed" && State != "Owner")) ) SUSPEND = ( $(KeyboardBusy) || \ ( (CpuBusyTime > 2 * $(MINUTE)) \ && $(ActivationTimer) > 90 ) ) CONTINUE = ( $(CPUIdle) && ($(ActivityTimer) > 10) \ && (KeyboardIdle > $(ContinueIdleTime)) ) PREEMPT = ( ((Activity == "Suspended") && \ ($(ActivityTimer) > $(MaxSuspendTime))) \ || (SUSPEND && (WANT_SUSPEND == False)) ) MaxJobRetirementTime = 0 KILL = $(ActivityTimer) > $(MaxVacateTime) PERIODIC_CHECKPOINT = ( (ImageSize < 60000) && \ ($(LastCkpt) > (6 * $(HOUR))) ) || \ ( $(LastCkpt) > (12 * $(HOUR)) )
This example shows how the default macros can be used to set up a machine for running test jobs from a specific user. Suppose we want the machine to behave normally, except if user coltrane submits a job. In that case, we want that job to start regardless of what is happening on the machine. We do not want the job suspended, vacated or killed. This is reasonable if we know coltrane is submitting very short running programs for testing purposes. The jobs should be executed right away. This works with any machine (or the whole pool, for that matter) by adding the following 5 expressions to the existing configuration:
START = ($(START)) || Owner == "coltrane" SUSPEND = ($(SUSPEND)) && Owner != "coltrane" CONTINUE = $(CONTINUE) PREEMPT = ($(PREEMPT)) && Owner != "coltrane" KILL = $(KILL)Notice that there is nothing special in either the CONTINUE or KILL expressions. If Coltrane's jobs never suspend, they never look at CONTINUE. Similarly, if they never preempt, they never look at KILL.
Condor can be configured to only run jobs at certain times of the day. In general, we discourage configuring a system like this, since you can often get lots of good cycles out of machines, even when their owners say ``I'm always using my machine during the day.'' However, if you submit mostly vanilla jobs or other jobs that cannot checkpoint, it might be a good idea to only allow the jobs to run when you know the machines will be idle and when they will not be interrupted.
To configure this kind of policy, you should use the ClockMin and ClockDay attributes, defined in section 3.5.1 on ``Startd ClassAd Attributes''. These are special attributes which are automatically inserted by the condor_ startd into its ClassAd, so you can always reference them in your policy expressions. ClockMin defines the number of minutes that have passed since midnight. For example, 8:00am is 8 hours after midnight, or 8 * 60 minutes, or 480. 5:00pm is 17 hours after midnight, or 17 * 60, or 1020. ClockDay defines the day of the week, Sunday = 0, Monday = 1, and so on.
To make the policy expressions easy to read, we recommend using macros to define the time periods when you want jobs to run or not run. For example, assume regular ``work hours'' at your site are from 8:00am until 5:00pm, Monday through Friday:
WorkHours = ( (ClockMin >= 480 && ClockMin < 1020) && \ (ClockDay > 0 && ClockDay < 6) ) AfterHours = ( (ClockMin < 480 || ClockMin >= 1020) || \ (ClockDay == 0 || ClockDay == 6) )
Of course, you can fine-tune these settings by changing the definition of AfterHours and WorkHours for your site.
Assuming you are using the default policy expressions discussed above, there are only a few minor changes required to force Condor jobs to stay off of your machines during work hours:
# Only start jobs after hours. START = $(AfterHours) && $(CPUIdle) && KeyboardIdle > $(StartIdleTime) # Consider the machine busy during work hours, or if the keyboard or # CPU are busy. MachineBusy = ( $(WorkHours) || $(CPUBusy) || $(KeyboardBusy) )
By default, the MachineBusy macro is used to define the SUSPEND and PREEMPT expressions. If you have changed these expressions at your site, you will need to add $(WorkHours) to your SUSPEND and PREEMPT expressions as appropriate.
Depending on your site, you might also want to avoid suspending jobs during work hours, so that in the morning, if a job is running, it will be immediately preempted, instead of being suspended for some length of time:
WANT_SUSPEND = $(AfterHours)
Suppose you have two classes of machines in your pool: desktop machines and dedicated cluster machines. In this case, you might not want keyboard activity to have any effect on the dedicated machines. For example, when you log into these machines to debug some problem, you probably do not want a running job to suddenly be killed. Desktop machines, on the other hand, should do whatever is necessary to remain responsive to the user.
There are many ways to achieve the desired behavior. One way is to make a standard desktop policy and a standard non-desktop policy and to copy the desired one into the local configuration file for each machine. Another way is to define one standard policy (in condor_ config) with a simple toggle that can be set in the local configuration file. The following example illustrates the latter approach.
For ease of use, an entire policy is included in this example. Some of the expressions are just the usual default settings.
# If "IsDesktop" is configured, make it an attribute of the machine ClassAd. STARTD_EXPRS = IsDesktop # Only consider starting jobs if: # 1) the load average is low enough OR the machine is currently # running a Condor job # 2) AND the user is not active (if a desktop) START = ( ($(CPUIdle) || (State != "Unclaimed" && State != "Owner")) \ && (IsDesktop =!= True || (KeyboardIdle > $(StartIdleTime))) ) # Suspend (instead of vacating/killing) for the following cases: WANT_SUSPEND = ( $(SmallJob) || $(JustCpu) || $(IsPVM) \ || $(IsVanilla) ) # When preempting, vacate (instead of killing) in the following cases: WANT_VACATE = ( $(ActivationTimer) > 10 * $(MINUTE) \ || $(IsPVM) || $(IsVanilla) ) # Suspend jobs if: # 1) The CPU has been busy for more than 2 minutes, AND # 2) the job has been running for more than 90 seconds # 3) OR suspend if this is a desktop and the user is active SUSPEND = ( ((CpuBusyTime > 2 * $(MINUTE)) && ($(ActivationTimer) > 90)) \ || ( IsDesktop =?= True && $(KeyboardBusy) ) ) # Continue jobs if: # 1) the CPU is idle, AND # 2) we've been suspended more than 5 minutes AND # 3) the keyboard has been idle for long enough (if this is a desktop) CONTINUE = ( $(CPUIdle) && ($(ActivityTimer) > 300) \ && (IsDesktop =!= True || (KeyboardIdle > $(ContinueIdleTime))) ) # Preempt jobs if: # 1) The job is suspended and has been suspended longer than we want # 2) OR, we don't want to suspend this job, but the conditions to # suspend jobs have been met (someone is using the machine) PREEMPT = ( ((Activity == "Suspended") && \ ($(ActivityTimer) > $(MaxSuspendTime))) \ || (SUSPEND && (WANT_SUSPEND == False)) ) # Replace 0 in the following expression with whatever amount of # retirement time you want dedicated machines to provide. The other part # of the expression forces the whole expression to 0 on desktop # machines. MaxJobRetirementTime = (IsDesktop =!= True) * 0 # Kill jobs if they have taken too long to vacate gracefully KILL = $(ActivityTimer) > $(MaxVacateTime)
With this policy in condor_ config, the local configuration files for desktops can be easily configured with the following line:
IsDesktop = True
In all other cases, the default policy described above will ignore keyboard activity.
Preemption can result in jobs being killed by Condor. When this happens, the jobs remain in the queue and will be automatically rescheduled. We highly recommend designing jobs that work well in this environment, rather than simply disabling preemption.
Planning for preemption makes jobs more robust in the face of other sources of failure. One way to live happily with preemption is to use Condor's standard universe, which provides the ability to produce checkpoints. If a job is incompatible with the requirements of standard universe, the job can still gracefully shutdown and restart by intercepting the soft kill signal.
All that being said, there may be cases where it is appropriate to force Condor to never kill jobs within some upper time limit. This can be achieved with the following configuration policy:
# When we want to kick a job off, let it run uninterrupted for # up to 2 days before forcing it to vacate. MAXJOBRETIREMENTTIME = $(HOUR) * 24 * 2
Construction of this expression may be more complicated. For example, it could provide a different retirement time to different users or different types of jobs. Also be aware that the job may come with its own definition of MaxJobRetirementTime, but this may only cause less retirement time to be used, never more than what the machine offers.
The longer the retirement time that is given, the slower reallocation of resources in the pool can become if there are long-running jobs. However, by preventing jobs from being killed, you may decrease the number of cycles that are wasted on non-checkpointable jobs that are killed.
Note that the use of MAXJOBRETIREMENTTIME limits the killing of jobs, but it does not prevent the preemption of resource claims. Therefore, it is technically not a way of disabling preemption, but simply a way of forcing preempting claims to wait until an existing job finishes or runs out of time. In other words, it limits the preemption of jobs but not the preemption of claims.
Limiting the preemption of jobs is often more desirable than limiting the preemption of resource claims. The following configuration limits the preemptions of resource claims:
#Disable preemption by machine activity. PREEMPT = False #Disable preemption by user priority. PREEMPTION_REQUIREMENTS = False #Disable preemption by machine RANK by ranking all jobs equally. RANK = 0 #Since we are disabling claim preemption, we # may as well optimize negotiation for this case: NEGOTIATOR_CONSIDER_PREEMPTION = False
Be aware of the consequences of this policy. Without any preemption of resource claims, once the condor_ negotiator gives the condor_ schedd a match to a machine, the condor_ schedd may hold onto this claim indefinitely, as long as the user keeps supplying more jobs to run. If this is not desired, force claims to be retired after some amount of time using CLAIM_WORKLIFE . This enforces a time limit, beyond which no new jobs may be started on an existing claim; therefore the condor_ schedd daemon is forced to go back to the condor_ negotiator to request a new match, if there is still more work to do.
Also be aware that in all versions of Condor prior to 6.8.1, it is not advisable to set NEGOTIATOR_CONSIDER_PREEMPTION to False, because of a bug that can lead to some machines never being matched to jobs.
Condor may be configured with a policy that allows these potentially evicted jobs to be suspended instead. The configuration introduces the notion of a virtual machine, even in the case of a single, physical CPU. The policy utilizes two virtual machines, one (called VM1 in the example) that runs jobs following the normal policy implemented for all machines in the pool. The second virtual machine (called VM2 in the example) is set to run jobs whenever it can, and suspend these jobs when VM1 is utilized.
Section 3.3.10 contains details of the STARTD_VM_EXPRS configuration macro, utilized in this policy example. The configuration, with comments to illustrate the intended policy will appear much like this example.
# Lie to Condor, to achieve 2 virtual machines with only a single CPU NUM_CPUS = 2 # VM1 is the high-prio VM, while VM2 is the background VM... START = (VirtualMachineID == 1) && $(VM1_START) || \ (VirtualMachineID == 2) && $(VM2_START) # Only start jobs on VM1 if the job is marked as a high-priority job VM1_START = (TARGET.IsHighPrioJob =?= TRUE) # Only start jobs on VM2 if there is no job on VM1, and if the machine is # otherwise idle... NOTE: the "Busy" activity is only in the Claimed # state, and only when there is an active job, so that is good enough # for our needs... VM2_START = ( (vm1_Activity != "Busy") && \ (KeyboardIdle > $(StartIdleTime)) && \ ($(CPUIdle) || (State != "Unclaimed" && State != "Owner")) ) # Only suspend jobs on VM2. Suspend if there is keyboard activity or # if a job starts on VM1... SUSPEND = (VirtualMachineID == 2) && \ ( (vm1_Activity == "Busy") || ($(KeyboardBusy)) ) # Since SUSPEND can only be TRUE on VM2, we can write the # CONTINUE expression specifically for VM2: CONTINUE = (KeyboardIdle > $(ContinueIdleTime)) && \ (vm1_Activity != "Busy")
This section describes how the current policy expressions differ from the policy expressions in previous versions of Condor. If you have never used Condor version 6.0 or earlier, or you never looked closely at the policy settings, skip this section.
In summary, there is no longer a VACATE expression, and the KILL expression is not evaluated while a machine is claimed. There is a PREEMPT expression which describes the conditions when a machine will move from the Claimed state to the Preempting state. Once a machine is transitioning into the Preempting state, the WANT_VACATE expression controls whether the job should be vacated with a checkpoint or directly killed. The KILL expression determines the transition from Preempting/Vacating to Preempting/Killing.
In previous versions of Condor, the KILL expression handled three distinct cases (the transitions from Claimed/Busy, Claimed/Suspended and Preempting/Vacating), and the VACATE expression handled two cases (the transitions from Claimed/Busy and Claimed/Suspended). In the current version of Condor, PREEMPT handles the same two cases as the previous VACATE expression, but the KILL expression handles one case. Very complex policies can now be specified using all of the default expressions, only tuning the WANT_VACATE and WANT_SUSPEND expressions. In previous versions, heavy use of the WANT_* expressions caused a complex KILL expression.