Charm++ parallel programming system

[François Tessier <francois.tessier AT inria.fr>]

Hmm... But do you know why it takes so much time for DummyLB (empty strategy, no migration) ? Here is an output example for DummyLB :

CharmLB> DummyLB: PE [0] step 0 starting at 51.107908 Memory: 95.636520 MB
CharmLB> DummyLB: PE [0] strategy starting at 51.295016
CharmLB> DummyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> DummyLB: PE [0] #Objects migrating: 0, LBMigrateMsg size: 0.02 MB
CharmLB> DummyLB: PE [0] strategy finished at 51.295266 duration 0.000250 s
CharmLB> DummyLB: PE [0] step 0 finished at 51.318621 duration 0.210713 s

CharmLB> DummyLB: PE [0] step 1 starting at 101.034147 Memory: 95.642670 MB
CharmLB> DummyLB: PE [0] strategy starting at 101.076371
CharmLB> DummyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> DummyLB: PE [0] #Objects migrating: 0, LBMigrateMsg size: 0.02 MB
CharmLB> DummyLB: PE [0] strategy finished at 101.076618 duration 0.000247 s
CharmLB> DummyLB: PE [0] step 1 finished at 101.082023 duration 0.047876 s

CharmLB> DummyLB: PE [0] step 2 starting at 150.783188 Memory: 95.643433 MB
CharmLB> DummyLB: PE [0] strategy starting at 152.562299
CharmLB> DummyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> DummyLB: PE [0] #Objects migrating: 0, LBMigrateMsg size: 0.02 MB
CharmLB> DummyLB: PE [0] strategy finished at 152.562545 duration 0.000246 s
CharmLB> DummyLB: PE [0] step 2 finished at 152.564820 duration 1.781632 s

CharmLB> DummyLB: PE [0] step 3 starting at 202.247956 Memory: 95.646362 MB
CharmLB> DummyLB: PE [0] strategy starting at 222.900881
CharmLB> DummyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> DummyLB: PE [0] #Objects migrating: 0, LBMigrateMsg size: 0.02 MB
CharmLB> DummyLB: PE [0] strategy finished at 222.901126 duration 0.000245 s
CharmLB> DummyLB: PE [0] step 3 finished at 222.903290 duration 20.655334 s

CharmLB> DummyLB: PE [0] step 4 starting at 272.647714 Memory: 95.647415 MB
CharmLB> DummyLB: PE [0] strategy starting at 299.725767
CharmLB> DummyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> DummyLB: PE [0] #Objects migrating: 0, LBMigrateMsg size: 0.02 MB
CharmLB> DummyLB: PE [0] strategy finished at 299.726014 duration 0.000247 s
CharmLB> DummyLB: PE [0] step 4 finished at 299.728286 duration 27.080572 s

CharmLB> DummyLB: PE [0] step 5 starting at 349.445116 Memory: 95.649063 MB
CharmLB> DummyLB: PE [0] strategy starting at 383.167981
CharmLB> DummyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> DummyLB: PE [0] #Objects migrating: 0, LBMigrateMsg size: 0.02 MB
CharmLB> DummyLB: PE [0] strategy finished at 383.168234 duration 0.000253 s
CharmLB> DummyLB: PE [0] step 5 finished at 383.170581 duration 33.725465 s

CharmLB> DummyLB: PE [0] step 6 starting at 432.921201 Memory: 95.649063 MB
CharmLB> DummyLB: PE [0] strategy starting at 471.529260
CharmLB> DummyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> DummyLB: PE [0] #Objects migrating: 0, LBMigrateMsg size: 0.02 MB
CharmLB> DummyLB: PE [0] strategy finished at 471.529507 duration 0.000247 s
CharmLB> DummyLB: PE [0] step 6 finished at 471.531730 duration 38.610529 s

Dr. François TESSIER
University of Bordeaux
Inria - TADaaM Team
Tel : 0033524574152
francois.tessier AT inria.fr
http://runtime.bordeaux.inria.fr/ftessier/
PGP 0x8096B5FA

On 15/06/2015 16:19, Harshitha Menon wrote:

First duration is for the strategy part. The second one is the total time taken by load balancing which includes stats collection, strategy and the migration. The time when strategy starts is given in the output (51.55 in the case of Greedy).

Can you provide me with the output of TreeMatchLB to help understand the timing.

Thanks
Harshitha

On Jun 15, 2015 6:35 AM, "François Tessier" <francois.tessier AT inria.fr> wrote:
Hello,

I'm looking for some explanations about the measures of time provided by
Charm++ when we call a load balancer. To give an example, I ran some
experiments on a Charm++ application with three load balancers :
DummyLB, GreedyLB and TreeMatchLB. At each LB step, Charm++ gives as
output something like this :

CharmLB> GreedyLB: PE [0] step 0 starting at 51.370771 Memory: 95.647156 MB
CharmLB> GreedyLB: PE [0] strategy starting at 51.553916
[...]
CharmLB> GreedyLB: PE [0] Memory: LBManager: 920 KB CentralLB: 2971 KB
CharmLB> GreedyLB: PE [0] #Objects migrating: 8188, LBMigrateMsg size:
0.52 MB
CharmLB> GreedyLB: PE [0] strategy finished at 52.032098 duration 0.478182 s
CharmLB> GreedyLB: PE [0] step 0 finished at 54.616962 duration 3.246191 s

My question is about the two durations given at the end of the LB step
(duration 1 : 0.478182 s / duration 2 : 3.246191 s). With my
application, the load balancer is called 7 times during the execution.
If I add these durations up :

Duration 1 (strategy finished at... duration ...) :
    GreedyLB      : 2.859333 s
    DummyLB     : 0.001735 s
    TreeMatchLB : 6.762819 s

Duration 2 (step n finished at... duration ...) :
    GreedyLB      : 36.826210 s
    DummyLB     : 122.112121 s
    TreeMatchLB : 136.887350 s

If I look at my own measures in TreeMatchLB, the first duration seems to
correspond to the execution time of the load balancer. Is it true ? What
is the second duration ? How could we explain this 136 seconds for
TreeMatchLB ?

Thanks for your help,

François

--
Dr. François TESSIER
University of Bordeaux
Inria - TADaaM Team
Tel : 0033524574152
francois.tessier AT inria.fr
http://runtime.bordeaux.inria.fr/ftessier/
PGP 0x8096B5FA

Attachment: signature.asc
Description: OpenPGP digital signature