Changes between Version 23 and Version 24 of GpuWorkFetch

Timestamp:

Jan 25, 2009, 8:32:27 PM (15 years ago)

Author:

davea

Comment:

--

GpuWorkFetch

@@ -7 +7 @@
  * If there's a CPU shortfall, request work from the project with highest long-term debt (LTD).
 
-The scheduler request has a scalar "work_req_seconds"
+The scheduler request has a single "work_req_seconds"
 indicating the total duration of jobs being requested.
 
-This policy has some problems.  First:
+This policy has some problems:
 
  * There's no way for the client to say "I have N idle CPUs; send me enough jobs to use them all".
@@ -38 +38 @@
 == Terminology ==
 
-New abstraction: '''processing resource''' or PRSC.
-The CPU is a PRSC and each coprocessor type is a PRSC.
+New abstraction: '''processing resource type''' or PRT.
+CPU and each coprocessor type are PRTs.
 
 A job sent to a client is associated with an app version,
@@ -54 +54 @@
 New fields in the scheduler request message:
 
- '''double cpu_req_seconds''':: number of CPU seconds requested
- '''double cpu_req_ninstances''':: send enough jobs to occupy this many CPUs
+ '''double cpu_req_secs''':: number of CPU seconds requested
+ '''double cpu_req_instances''':: send enough jobs to occupy this many CPUs
 
 And for each coprocessor type:
 
- '''double req_seconds''':: number of instance-seconds requested
- '''double req_ninstances''':: send enough jobs to occupy this many instances
+ '''double req_secs''':: number of instance-seconds requested
+ '''double req_instances''':: send enough jobs to occupy this many instances
 
 For compatibility with old servers, the message still has '''work_req_seconds''',