Re: [google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
The app engine imposes quotas on various parameters to ensure that your app doesn't hijack the system resources of other's apps and other's apps don't hijack your app's systems resources. http://code.google.com/appengine/docs/quotas.html This link will help you with more info on that. Thanks, - Madhu On Thu, Dec 23, 2010 at 12:51 PM, Matija matija.jerko...@gmail.com wrote: GAE team !? More info. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.comgoogle-appengine%2bunsubscr...@googlegroups.com . 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 google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
Re: [google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
I understand quota system, but I don't know exact parameters that trigger additional latency in statement: 'Applications that are heavily CPU-bound may also incur some additional latency in order to efficiently share resources with other applications on the same servers.' Also I don't know how this latency is imposed. I would like to know them before they are imposed or if they will be imposed. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
Re: [google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
The text for this quota warning originated in 2008 or begin 2009. At that time there was a more strict quota on CPU usage. Now you use the warning to locate requests that might be up for an improvement. Inspect the request with app-stat and rethink the code used. 2010/12/23 Matija matija.jerko...@gmail.com: I understand quota system, but I don't know exact parameters that trigger additional latency in statement: 'Applications that are heavily CPU-bound may also incur some additional latency in order to efficiently share resources with other applications on the same servers.' Also I don't know how this latency is imposed. I would like to know them before they are imposed or if they will be imposed. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
Re: [google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
GAE team !? More info. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
[google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
Sub-400 ms as optimal is for request latency time and I agree with them. You will probably get new instances if your average latency is under 1000 ms, but if it is under 400 ms you would get new instances at faster rate. Maybe. I don't have problem with request latency. Actually I don't have problem with GAE at all. I would like to know a little bit more about CPU usage and scaling up new instances. What is meaning of '...may also incur some additional latency in order to efficiently share resources...' ? How did they implemented this additional latency? With no addition of new instances after some limit or some penalty latency (throttle_code?) or something else ?! -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
[google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
Are you sure about this ? I have noticed that I did indeed get new additional instance, but are their automatic scaling algorithm so simple that uses only average latency time. Maybe this is truth for first 10 instances and then they add additional checking. Although they nowhere display average cpu_ms or api_cpu_ms per instance we can't be sure that they don't collect it and use it. Few bytes per instance and few mathematical operation. I would collect it. With low latency they are able to easily scale applications on application servers, but with high cpu/api cpu ms they are not able to easily scale datastore or any other api service. So we can't say that 160 ms request with 17000 cpu ms is nice scalable request, but is 160 ms request with 1700 cpu ms ok or not. So is there some good max cpu/api cpu policy that they didn't explained ? Or they did with this yellow and red coloring schema around 700/1000 ms cpu usage. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
Re: [google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
If I remember right, think the coloring scheme is left over from when there were other quotas in place. On Fri, Dec 17, 2010 at 03:09, Matija matija.jerko...@gmail.com wrote: Are you sure about this ? I have noticed that I did indeed get new additional instance, but are their automatic scaling algorithm so simple that uses only average latency time. Maybe this is truth for first 10 instances and then they add additional checking. Although they nowhere display average cpu_ms or api_cpu_ms per instance we can't be sure that they don't collect it and use it. Few bytes per instance and few mathematical operation. I would collect it. With low latency they are able to easily scale applications on application servers, but with high cpu/api cpu ms they are not able to easily scale datastore or any other api service. So we can't say that 160 ms request with 17000 cpu ms is nice scalable request, but is 160 ms request with 1700 cpu ms ok or not. So is there some good max cpu/api cpu policy that they didn't explained ? Or they did with this yellow and red coloring schema around 700/1000 ms cpu usage. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. 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 google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
Re: [google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
Yes, I remember this. But after removal of 'high cpu quota' there was no (spoken) penalty for over 1000 ms request latency (beside bad user UI experience) and now it is inability to get new instances. If you look at Quotas and Limitshttp://code.google.com/appengine/docs/java/runtime.html#Quotas_and_Limits there is statement: 'Applications that are heavily CPU-bound may also incur some additional latency in order to efficiently share resources with other applications on the same servers.' So what are heavily CPU-bound applications. Currently only information that we have are yellow and red warnings. I hope that somebody from GAE team will give us more information. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
[google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
I had a thread saved where Nick or someone else on the team confirmed the 1000 ms boundary, however it seems my saved thread collection has grown ridiculously big and I can't find it. I did find this thread discussing it however: http://groups.google.com/group/google-appengine/browse_thread/thread/a5d1aeb359ff03ab -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
[google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
Somebody (Ikai L if I remember well) even mentioned sub-400ms as optimal number. I agree we need more light on this. For the complete (almost) story of those numbers you can take a look at my post here: http://gaengine.blogspot.com/2010/09/app-engine-scalability-issues.html Happy coding :-) Nick On Dec 17, 10:43 pm, Darien Caldwell darien.caldw...@gmail.com wrote: I had a thread saved where Nick or someone else on the team confirmed the 1000 ms boundary, however it seems my saved thread collection has grown ridiculously big and I can't find it. I did find this thread discussing it however: http://groups.google.com/group/google-appengine/browse_thread/thread/... -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
[google-appengine] Re: This request used a high amount of CPU and may soon exceed its quota - 2010/2011
The only penalty is that requests that take more than 1000 mS will not benefit from automatic scaling. And you'll use more quota, obviously. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To post to this group, send email to google-appeng...@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
