harshitgupta1337
In improv version, No energy consumption infomation
Hello, in the previous version, by running the VRGameFog.java, we can get the energy consumption details from the simulation results. In the newest improv and mobility versions, the energy consumed is no longer available. Is there any way to modify the PhyTopoExample.java to get the energy information?
Due to change in scheduling policy in improv branch, energy calculation using old method has been disabled. I am working on reporting correct energy values.
On Fri, Jun 16, 2017 at 11:15 AM, yawenluo16 [email protected] wrote:
Hello, in the previous version, by running the VRGameFog.java, we can get the energy consumption details from the simulation results. In the newest improv and mobility versions, the energy consumed is no longer available. Is there any way to modify the PhyTopoExample.java to get the energy information?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/harshitgupta1337/fogsim/issues/8, or mute the thread https://github.com/notifications/unsubscribe-auth/ADjC4BoETGYsKeQ1ZTeLePFB16rxSM9tks5sEhZtgaJpZM4N8DBZ .
-- With regards,
HARSHIT GUPTA PhD Student Embedded Pervasive Laboratory Georgia Institute of Technology Atlanta, GA 30332. +1 (404) 512-9792 [email protected]
Well, I just found that in the master version, when I run the DCNSFog.java, yeah I can get the energy consumed information, but when enables the log, go into the details of the console, I think there is a little problem: at any time, the utilization of any host is always 0%. Which means the energy consumption calculated is just the "idle power* execution time". It's inconsistent with the LinearPowerModel.
Yes that is because of the way MIPS are allocated to Vms in master branch. I was not able to fix the zero utilization issue, so had my own function for calculating utilization and then energy consumption using given power model.
On Jun 16, 2017 9:47 PM, "yawenluo16" [email protected] wrote:
Well, I just found that in the master version, when I run the DCNSFog.java, yeah I can get the energy consumed information, but when enables the log, go into the details of the console, I think there is a little problem: at any time, the utilization of any host is always 0%. Which means the energy consumption calculated is just the "idle power* execution time". It's inconsistent with the LinearPowerModel. [image: capture] https://user-images.githubusercontent.com/29474686/27235143-463ee69c-5285-11e7-88f6-4dec8398495a.PNG
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/harshitgupta1337/fogsim/issues/8#issuecomment-309069339, or mute the thread https://github.com/notifications/unsubscribe-auth/ADjC4JbdBIjoSYCpWVQMcaZRgKluaLSLks5sEqqLgaJpZM4N8DBZ .
In the above pic, when I am processing on device d-0 and d-1 (for an EEG_Transmission_Time > 6.2 and no.of.mobilesperdept >5) the energy consumption on device d-0 and d-1 are giving the minimum power. Is there any threshold value? could anyone please help in this?
Hello yawenluo16, I need some information on calculation of an energy consumption.
1)Could you please tell me, why did we multiply 16 with min and max power in cloud? ( I mean 16(103) and 16(83.25) what exactly mean 16 here?
The way why it multiplies 16, I think there is just an assumption. You can look up some materials for what's the power of the general cloud server or make your own hypothesis according to your application.
On Mon, Nov 13, 2017 at 7:01 PM, Santoshreddy709 [email protected] wrote:
Hello yawenluo16, I need some information on calculation of an energy consumption.
1)Could you please tell me, why did we multiply 16 with min and max power in cloud? ( I mean 16(103) and 16(83.25) what exactly mean 16 here?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/harshitgupta1337/fogsim/issues/8#issuecomment-344110882, or mute the thread https://github.com/notifications/unsubscribe-auth/AcG_fn-bMIWuOurN3UAHz276io61wNDtks5s2OZmgaJpZM4N8DBZ .
Do you mean what's the energy consumption's minimum value? Based on my understanding, the energy consumed is decided by the minimum power of the device, the processing time, as well as the amounts of loads. When the loads are light, the consumption is usually the min_power * time. When it's overload, will be max_power * time.
Above it's just my view, you may ask the author for help!
On Mon, Nov 13, 2017 at 6:59 PM, Santoshreddy709 [email protected] wrote:
[image: image] https://user-images.githubusercontent.com/27026267/32757289-7496022e-c8a4-11e7-8434-f06bdccb280f.png
In the above pic, when I am processing on device d-0 and d-1 (for an EEG_Transmission_Time > 6.2 and no.of.mobilesperdept >5) the energy consumption on device d-0 and d-1 are giving the minimum power. Is there any threshold value? could anyone please help in this?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/harshitgupta1337/fogsim/issues/8#issuecomment-344110442, or mute the thread https://github.com/notifications/unsubscribe-auth/AcG_fooF3EcvXLX4RMRfSMfF9VC4M3fZks5s2OXdgaJpZM4N8DBZ .
-- Yawen Luo PhD Candidate Electrical and Computer Engineering Department
University of [email protected] [email protected]
W345 Engr. Bldg. 2
Houston, Texas 77204