Beautiful! I really needed this--thanks.
But, "Delegate" is an interface, so I think the ProfilingDelegate
needs to look something like this:
import com.google.apphosting.api.ApiProxy.Delegate;
import com.google.apphosting.api.ApiProxy.Environment;
import com.google.apphosting.api.ApiProxy.LogRecord;
class ProfilingDelegate implements Delegate {
Delegate parent;
public ProfilingDelegate(Delegate parent) {
this.parent = parent;
}
public byte[] makeSyncCall(Environment env, String pkg, String
method, byte[] request) {
long start = System.nanoTime();
byte[] result = parent.makeSyncCall(env, pkg, method, request);
log.info(pkg + "." + method + ": " + System.nanoTime() - start);
return result;
}
public void log(Environment env, LogRecord logRec) {
parent.log(env, logRec);
}
}
On Thu, Aug 20, 2009 at 12:02 PM, Toby Reyelts<[email protected]> wrote:
> You'll need to profile two separate things:
> 1) Compute-bound execution
> 2) Time spent in rpcs to our services
> 1) Compute-bound performance should be in the same rough ballpark on the
> dev_appserver as our production servers. You can use any standard Java
> profiler on your dev_appserer to measure this.
> 2) Time spent in services is going to vary wildly between production and the
> dev_appserver. For example, the datastore for the dev_appserver is just an
> in-memory store with periodic writebacks. The datastore for production is a
> distributed database. You can measure these times in production using a
> profiling ApiProxy.Delegate:
> class ProfilingDelegate extends Delegate {
> Delegate parent; public ProfilingDelegate(Delegate parent) { this.parent =
> parent; } public byte[] makeSyncCall(Environment env, String pkg, String
> method, byte[] request) { long start = System.nanoTime(); byte[] result =
> parent.makeSyncCall(env, pkg, method, request); log.log(INFO, pkg + “.” +
> method + “: “ + System.nanoTime() - start); return result; } }
> ApiProxy.setDelegate(new ProfilingDelegate(ApiProxy.getDelegate()));
> If you're feeling particularly aggressive, you can use the App Engine status
> site in conjunction with an ApiProxy.Delegate to come up with an estimation
> of how much time you'll spend in RPCs on the production and even insert
> corresponding sleeps into your dev_appserver.
> On Thu, Aug 20, 2009 at 5:17 AM, mar_novice <[email protected]> wrote:
>>
>> is there a way to do profiling with google app engine java?
>>
>
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Google App Engine for Java" 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-java?hl=en
-~----------~----~----~----~------~----~------~--~---
