On Wed, Sep 7, 2011 at 8:56 AM, Tammo Freese <[email protected]> wrote: > Hi Jon, > > > again, thanks for your answer, and sorry for bothering you with new > questions. This would be so much easier in face-to-face conversation > than with email. If you would like a Google hangout, just drop me an > email off-list. > > On Sep 7, 4:49 pm, Jon McAlister <[email protected]> wrote: >> On Wed, Sep 7, 2011 at 7:20 AM, Tammo Freese <[email protected]> wrote: >> > so total-instances is the maximum of active-instances over the last 15 >> > minutes? >> >> Not really, no. > > So what is total-instances then? I know, that sounds like a dumb > question. At first assumed that in second x, it would simply be the > number of instances (active+idle) in that second (total instances as I > understand it). But in one of your posts, you wrote "total-instances > refers to the blue line on the graph, and is computed according to the > +15-minutes-since-last-request formula".
Right. The way the total-instances-rate (the blue line) is computed is from all running instances that have received a request at any point in the last 15 minutes. This however is not equal to the line you proposed, the line which is the maximum value of active-instances-rate over the last 15 minutes. Look at any app's instances graph and you will be able to see this visually right away. > > >> > Let's say I have 4 active instances for 5 minutes, then no traffic for >> > the rest of the day, max-idle-instances set to 1. >> > 1) 35 minutes will be billed regardless when the scheduler decides to >> > kill idle instances, right? >> >> Yes. >> >> > 2) But 1 instance would be idle for at least 15 minutes, right? >> >> Maybe. > > [...] > >> > Let's say I have 4 active instances for 5 minutes, then no traffic for >> > the rest of the day, max-idle-instances set to 4. >> > 1) 80 minutes will be billed regardless when the scheduler decides to >> > kill idle instances, right? >> >> Yes. >> >> > 2) But 4 instances would be idle for at least 15 minutes, right? >> >> Maybe. > > So setting max-idle-instances to 4 leads to being reliably billed > more, but not having any reliable way of measuring the benefit? The benefit should be visible in terms of average serving latency and reliability. There should be 4 idle instances running over the time period you proposed, but there is not a hard guarantee of this. > > > Thanks, > > Tammo > > -- > You received this message because you are subscribed to the Google Groups > "Google App Engine" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]. > For more options, visit this group at > http://groups.google.com/group/google-appengine?hl=en. > > -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
