I see. Thanks, Robert!
On May 5, 11:14 pm, Robert Kluin <[email protected]> wrote: > Hi Albert, > It is probably the best metric you can easily see. But keep in > mind, that number also includes tasks. So, if your app makes use of > lots of long-running tasks it could bias the number up (or, > conversely, lots of super fast tasks could bias it down). Also, if > your app's loading time is very long relative to the typical time it > takes to serve a request the scheduler seems to prefer waiting for an > existing instance to serve the request. > > You'll know your app is scaling if you get new instances spun-up as > traffic increases. If you start seeing lots of pending times in your > logs, and/or a sudden spike in latency that could indicate an issue. > > Robert > > > > > > > > On Thu, May 5, 2011 at 08:03, Albert <[email protected]> wrote: > > Thanks! > > > I took a look at what you told me, and I think I prefer the instances > > section. At least I only have to look at a few numbers as opposed to > > summarizing all the request logs. > > > To everyone else, is the instances section a good place to find > > information on whether an app is qualified to autoscale or not? > > > Thanks! > > > On May 5, 6:05 pm, Fabs <[email protected]> wrote: > >> Hello > > >> If you look in the logs section in the dashboard then select "All > >> requests" you can see the request latency of each request. If you go > >> to the instances section of the dashboard you can see the average > >> request latency of each instance of your app. I wasn't aware of a > >> limit on user-facing request times for scaling, but perhaps app engine > >> is less willing to scale apps that respond slowly. At any rate, 800ms > >> is a long time for most common web requests. If you're doing more than > >> a second's worth of work such as complex datastore operations I would > >> at least consider using the task queue. > > >> Regards. > > >> On May 5, 7:08 pm, Albert <[email protected]> wrote: > > >> > Hi! > > >> > Is it possible for me to know if my app is "qualified" to autoscale > >> > just by looking at the values being displayed in the dashboard? > > >> > It currently gives info on CPU and API CPU average request times. > >> > However, please correct me if I'm wrong, I remember that the measure > >> > for autoscaling is that your user facing requests should stay below > >> > 800 - 1000ms in response time (even if the CPU/API CPU is above > >> > 1000ms). > > >> > If this is the case, can I tell just by looking at the dashboard if my > >> > app is ok to autoscale or not? Or do I have to dig in to the logs and > >> > use appstats to manually profile my app? > > >> > Thanks! > > > -- > > 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 > > athttp://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.
