Re: [Autoscaling]Instance count rounding algorithm

2014-11-26 Thread Lahiru Sandaruwan
I have done the changes so that we can specify this value in Autoscaling policy. Thanks. On Sat, Nov 8, 2014 at 8:35 PM, Rajkumar Rajaratnam wrote: > > > On Sat, Nov 8, 2014 at 8:32 PM, Lahiru Sandaruwan > wrote: > >> >> >> On Sat, Nov 8, 2014 at 8:26 PM, Rajkumar Rajaratnam >> wrote: >> >>>

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Lahiru Sandaruwan
On Sat, Nov 8, 2014 at 9:06 PM, Reka Thirunavukkarasu wrote: > Hi > > On Sat, Nov 8, 2014 at 7:57 PM, Lahiru Sandaruwan > wrote: > >> Hi all, >> >> I'm working on $subject. When we predict the instance count from >> statistics and dependency based ratio in application, we will get numbers >> wit

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Reka Thirunavukkarasu
Hi On Sat, Nov 8, 2014 at 7:57 PM, Lahiru Sandaruwan wrote: > Hi all, > > I'm working on $subject. When we predict the instance count from > statistics and dependency based ratio in application, we will get numbers > with fraction. E.g. 1.2, 3.9, 4.5. Therefore we have to round the instance > co

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Rajkumar Rajaratnam
On Sat, Nov 8, 2014 at 8:32 PM, Lahiru Sandaruwan wrote: > > > On Sat, Nov 8, 2014 at 8:26 PM, Rajkumar Rajaratnam > wrote: > >> Hi Lahiru, >> >> Are you going to make all three configurable via autoscaler xml? or How >> you are going to do? >> >> > IMO it should be a part of Autoscaling policy.

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Rajkumar Rajaratnam
On Sat, Nov 8, 2014 at 8:24 PM, Lahiru Sandaruwan wrote: > > > On Sat, Nov 8, 2014 at 8:14 PM, Rajkumar Rajaratnam > wrote: > >> Oh, I think I am confused here. Ignore the previous mail. >> >> But #3 can cover all the scenario :) >> > > Yes. User can input 0 for #1 and 1 for #2. May be we can ju

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Lahiru Sandaruwan
On Sat, Nov 8, 2014 at 8:26 PM, Rajkumar Rajaratnam wrote: > Hi Lahiru, > > Are you going to make all three configurable via autoscaler xml? or How > you are going to do? > > IMO it should be a part of Autoscaling policy. > Thanks. > > On Sat, Nov 8, 2014 at 8:22 PM, Lahiru Sandaruwan > wrote:

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Rajkumar Rajaratnam
Hi Lahiru, Are you going to make all three configurable via autoscaler xml? or How you are going to do? Thanks. On Sat, Nov 8, 2014 at 8:22 PM, Lahiru Sandaruwan wrote: > > > On Sat, Nov 8, 2014 at 8:06 PM, Nirmal Fernando > wrote: > >> Why not simply round the value to nearest integer? >> ht

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Lahiru Sandaruwan
On Sat, Nov 8, 2014 at 8:14 PM, Rajkumar Rajaratnam wrote: > Oh, I think I am confused here. Ignore the previous mail. > > But #3 can cover all the scenario :) > Yes. User can input 0 for #1 and 1 for #2. May be we can just have one option to the users then. > > Thanks. > > On Sat, Nov 8, 2014

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Lahiru Sandaruwan
On Sat, Nov 8, 2014 at 8:06 PM, Nirmal Fernando wrote: > Why not simply round the value to nearest integer? > http://docs.oracle.com/javase/7/docs/api/java/lang/Math.html#round(float) > It is not the industry practice. AWS guys use a more complex algorithms. But we can start with suggested optio

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Rajkumar Rajaratnam
Oh, I think I am confused here. Ignore the previous mail. But #3 can cover all the scenario :) Thanks. On Sat, Nov 8, 2014 at 8:07 PM, Rajkumar Rajaratnam wrote: > Hi Lahiru, > > #3 includes #1 and #2. > > If user input < 0.5 implies #1 > If user input >= 0.5 implies #2 > > IMO, we can go with

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Nirmal Fernando
Why not simply round the value to nearest integer? http://docs.oracle.com/javase/7/docs/api/java/lang/Math.html#round(float) On Sat, Nov 8, 2014 at 3:28 PM, Lahiru Sandaruwan wrote: > /s/fives/gives > > On Sat, Nov 8, 2014 at 7:57 PM, Lahiru Sandaruwan > wrote: > >> Hi all, >> >> I'm working on

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Rajkumar Rajaratnam
Hi Lahiru, #3 includes #1 and #2. If user input < 0.5 implies #1 If user input >= 0.5 implies #2 IMO, we can go with only #3. wdyt? Thanks. On Sat, Nov 8, 2014 at 7:57 PM, Lahiru Sandaruwan wrote: > Hi all, > > I'm working on $subject. When we predict the instance count from > statistics an

Re: [Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Lahiru Sandaruwan
/s/fives/gives On Sat, Nov 8, 2014 at 7:57 PM, Lahiru Sandaruwan wrote: > Hi all, > > I'm working on $subject. When we predict the instance count from > statistics and dependency based ratio in application, we will get numbers > with fraction. E.g. 1.2, 3.9, 4.5. Therefore we have to round the i

[Autoscaling]Instance count rounding algorithm

2014-11-08 Thread Lahiru Sandaruwan
Hi all, I'm working on $subject. When we predict the instance count from statistics and dependency based ratio in application, we will get numbers with fraction. E.g. 1.2, 3.9, 4.5. Therefore we have to round the instance count for creating instances in IaaS. I have thought of three types to cho