We ran into some disk issues that have delayed my testing. We are not sure, but it might also have been causing the problem that we saw when running the large Lucene merges. I will send out another note once our disk problems are fixed. Lokesh
-----Original Message----- From: Otis Gospodnetic [mailto:[EMAIL PROTECTED] Sent: Monday, June 27, 2005 10:08 AM To: java-user@lucene.apache.org Subject: Re: issues building a large index Hi, Perhaps using hprof with cpu=samples may reveal more information about what CPU is doing. I think this is a valid use case. Otis --- Lokesh Bajaj <[EMAIL PROTECTED]> wrote: > Thanks for the idea. I have tried with both 512m and 1024m with the > same results. I also turned on verbose gc logging. Once I see this > slowdown, the gc is running only once every 12 minutes or so and > finishes in less than 0.005 seconds. > > Its basically stuck at 100% cpu usage, but doesn't seem to be doing > anything useful. Lokesh > > > > Daniel Naber <[EMAIL PROTECTED]> wrote: > On Saturday 25 June 2005 02:10, Lokesh Bajaj wrote: > > > 3] Does this seem like a JVM issue? Since its always pointing to a > > native method, I am not really sure what to look for or debug. > > Does you JVM have enough heap (e.g. -Xmx500M)? Java gets slow if it's > busy with garbage collection. > > Regards > Daniel --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
