Details
-
Type: Improvement
-
Status: Closed
-
Priority: Major
-
Resolution: Fixed
-
Affects Version/s: Current Version
-
Fix Version/s: None
-
Component/s: Evaluation Engine / Workers
-
Labels:None
Description
Motivation:
1) Now we have to predict how much RAM we should reserve for single jobs. If we choose 1G, then we are lowering our possible concurrency. If we choose 512M, then worker could potentially lack memory.
2) We have different configs for judging machines which vary only in concurrency.
What to do:
1) Replace concurrency in worker config with RAM capacity
2) OIOIOI should add to environ information how much RAM single job needs
3) Scheduler should operate on RAM capacities instead of concurrencies.
4) Split this task into smaller ones :)
Adnotation:
Don't forget about taking into account number of CPU cores. Currently used formula is min(# cpu cores * 1.5, (RAM minus a little bit for OS) / (max required mem per problem)). Obviously it is worth to discuss if and how this formula should be changed.
1) Now we have to predict how much RAM we should reserve for single jobs. If we choose 1G, then we are lowering our possible concurrency. If we choose 512M, then worker could potentially lack memory.
2) We have different configs for judging machines which vary only in concurrency.
What to do:
1) Replace concurrency in worker config with RAM capacity
2) OIOIOI should add to environ information how much RAM single job needs
3) Scheduler should operate on RAM capacities instead of concurrencies.
4) Split this task into smaller ones :)
Adnotation:
Don't forget about taking into account number of CPU cores. Currently used formula is min(# cpu cores * 1.5, (RAM minus a little bit for OS) / (max required mem per problem)). Obviously it is worth to discuss if and how this formula should be changed.
Activity
Artur Jamro
made changes -
Field | Original Value | New Value |
---|---|---|
Summary | Replace concurrency in worker with RAM capacity. | Replace worker concurrency with RAM capacity. |
Artur Jamro
made changes -
Description |
Motivation:
1) Now we have to predict how much RAM we should reserve for single jobs. If we choose 1G, then we are lowering our possible concurrency. If we choose 512M, then worker could potentially lack memory. 2) We have different configs for judging machines which vary only in concurrency. What to do: 1) Replace concurrency in worker config with RAM capacity 2) OIOIOI should add to environ information how much RAM single job needs 3) Scheduler should operate on RAM capacities instead of concurrencies. 4) Split this task into smaller ones :) |
Motivation:
1) Now we have to predict how much RAM we should reserve for single jobs. If we choose 1G, then we are lowering our possible concurrency. If we choose 512M, then worker could potentially lack memory. 2) We have different configs for judging machines which vary only in concurrency. What to do: 1) Replace concurrency in worker config with RAM capacity 2) OIOIOI should add to environ information how much RAM single job needs 3) Scheduler should operate on RAM capacities instead of concurrencies. 4) Split this task into smaller ones :) Adnotation: Don't forget about taking into account number of CPU cores. |
Artur Jamro
made changes -
Description |
Motivation:
1) Now we have to predict how much RAM we should reserve for single jobs. If we choose 1G, then we are lowering our possible concurrency. If we choose 512M, then worker could potentially lack memory. 2) We have different configs for judging machines which vary only in concurrency. What to do: 1) Replace concurrency in worker config with RAM capacity 2) OIOIOI should add to environ information how much RAM single job needs 3) Scheduler should operate on RAM capacities instead of concurrencies. 4) Split this task into smaller ones :) Adnotation: Don't forget about taking into account number of CPU cores. |
Motivation:
1) Now we have to predict how much RAM we should reserve for single jobs. If we choose 1G, then we are lowering our possible concurrency. If we choose 512M, then worker could potentially lack memory. 2) We have different configs for judging machines which vary only in concurrency. What to do: 1) Replace concurrency in worker config with RAM capacity 2) OIOIOI should add to environ information how much RAM single job needs 3) Scheduler should operate on RAM capacities instead of concurrencies. 4) Split this task into smaller ones :) Adnotation: Don't forget about taking into account number of CPU cores. Currently used formula is min(# cpu cores * 1.5, (RAM minus a little bit for OS) / (max required mem per problem)). Obviously it is worth to discuss if and how this formula should be changed. |
Pavel Senchanka
made changes -
Status | New [ 10000 ] | Resolved [ 5 ] |
Assignee | Szymon Acedański [ accek ] | |
Resolution | Fixed [ 1 ] |
Szymon Acedański
made changes -
Status | Resolved [ 5 ] | Closed [ 6 ] |
Transition | Time In Source Status | Execution Times | Last Executer | Last Execution Date | |||||
---|---|---|---|---|---|---|---|---|---|
|
387d 21h 6m | 1 | Pavel Senchanka | 2018-04-17 12:40 | |||||
|
343d 4h 59m | 1 | Szymon Acedański | 2019-03-26 16:40 |
https://gerrit.sio2project.mimuw.edu.pl/2960
SIO-2071Adapt sioworkersd to respect RAM limits.This commit is a large part of the implementation of
SIO-1928. It modifies sioworkersd (mainly the defaultPrioritizingScheduler) to recognize workers' total
RAM amounts and tasks' required RAM amount. The
core scheduling algorithm is kept the same, but
modified to account for RAM limits, which may have
some impact on the performance.
For these changes to be fully functional, it's also
necessary to implement
SIO-2070andSIO-2073.Change-Id: I625f7e7303bd5f69a96bde1443f8d9c38ef2ef03