[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[condor-users] Condor DAGMan performance question
- Date: Mon, 26 Apr 2004 01:08:41 -0700 (PDT)
- From: Yuhong Feng <april_continue@xxxxxxxxx>
- Subject: [condor-users] Condor DAGMan performance question
I am testing the performance of DAGMan by using the
Parent MCCF_A Child MCCF_B
Parent MCCF_B Child MCCF_C
Parent MCCF_C Child MCCF_D
Parent MCCF_D Child MCCF_E
and Job MCCF_A, MCCF_B, MCCF_C, MCCF_D, MCCF_E will
run on machine A, B, C, D, E respectively. Job
MCCF_B will use the result of MCCF_A, and MCCF_C
will use the result of MCCF_B and so on. the workload
is stable for each machine, so as the network status.
each job output volume is the same size. the output
volume will be increased.
However the result is :
y= [401 604 459 564 367 606 767 510 546 788 805]
x=[560000 2800000 5600000 8400000 11200000 14000000
16800000 19600000 22400000 25200000 28000000]
x is output data volume of Job, y is the execution
time of the whole DAG. As we can see, the execution
time is not increased corresponding to the data volume
increase. When data increase, this will call for more
data transmission time over the network.
only one Dagman job submited at a time.
Would you please give me ths possible reason? your
kind suggestions about regard to condor DAGMan
performance are highly appreciated.
Do you Yahoo!?
Yahoo! Photos: High-quality 4x6 digital prints for 25¢
Condor Support Information:
To Unsubscribe, send mail to majordomo@xxxxxxxxxxx with
unsubscribe condor-users <your_email_address>