Changes between Version 67 and Version 68 of ProjectOptions

Timestamp:

Jun 1, 2008, 3:05:10 PM (16 years ago)

Author:

davea

Comment:

--

ProjectOptions

@@ -12 +12 @@
 }}}
 
-== Scheduling: limiting work distribution ==
+== Scheduling: general ==
 
 {{{
@@ -63 +63 @@
 }}}
 If set, [AppFiltering user application selection] applies to [BetaTest beta test applications] as well as others.
-
-== Scheduling: array-based scheduling ==
 {{{
 <nowork_skip> 0|1 </nowork_skip>
 }}}
 If the scheduling server has no work, it replies to RPCs without doing any database access (e.g., without looking up the user or host record). This reduces DB load, but it fails to update preferences when users click on Update. Use it if your server DB is overloaded.
+
+== Scheduling: job-cache scheduling ==
+
+This is the default.
+The feeder (a daemon program) maintains a cache of jobs in shared memory.
+Instances of the scheduler get jobs from this cache,
+reducing their database access overhead.
+
 {{{
 <shmem_work_items>N</shmem_work_items>
 }}}
-The size of the shared-memory array of jobs.  Default is 100.
+The size of the job cache.  Default is 100 jobs.
 {{{
 <feeder_query_size>N</feeder_query_size>
@@ -96 +102 @@
 Results with priority at least 'reliable_on_priority' will be sent only to reliable hosts; increase priority of duplicate results by 'reliable_priority_on_over'; increase priority of duplicates caused by timeout (not error) by 'reliable_priority_on_over_except_error'.
 
+== Scheduling: matchmaker scheduling ==
+
+This is a variant of job-cache scheduling.
+The job selection policy is determined by a "score function";
+this highest-scoring jobs are sent to the client.
+
+{{{
+<matchmaker>0|1</matchmaker>
+}}}
+If set, enable matchmaker scheduling.
+{{{
+<mm_min_slots>N</mm_min_slots>
+<mm_max_slots>N</mm_max_slots>
+}}}
+Specify the min and max number of jobs to scan for a given client request.
+Defaults are 20 and 50.
+{{{
+<job_size_matching>0|1</job_size_matching>
+}}}
+If set, include a term in the score function that
+favors sending large jobs to fast hosts.
+To use this, you must run the census program as a periodic task
+to maintain statistics on the distribution of host speeds.
 
 == Scheduling: locality scheduling ==