Received: by 10.236.116.74 with SMTP id f50mr16310376yhh.33.1349210997071; Tue, 02 Oct 2012 13:49:57 -0700 (PDT) X-BeenThere: project-voldemort@googlegroups.com Received: by 10.236.155.36 with SMTP id i24ls1518377yhk.3.gmail; Tue, 02 Oct 2012 13:49:55 -0700 (PDT) Received: by 10.236.183.101 with SMTP id p65mr2655411yhm.20.1349210995568; Tue, 02 Oct 2012 13:49:55 -0700 (PDT) Date: Tue, 2 Oct 2012 13:49:54 -0700 (PDT) From: ctasada To: project-voldemort@googlegroups.com Message-Id: Subject: AdminClient fetchKeys performance MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="----=_Part_1481_4044202.1349210994555" ------=_Part_1481_4044202.1349210994555 Content-Type: multipart/alternative; boundary="----=_Part_1482_23184810.1349210994555" ------=_Part_1482_23184810.1349210994555 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Hi guys, What's the expected performance of the AdminClient.fetchKeys? I'm using it but the behaviour is a bit heretic. I get something like 10K keys every 1.5 seconds, which is good, but then suddenly, the streaming stops for 30 to 60 seconds. My suspicious is that the stops are caused by BDB reading, since some keys may not be cached. Is there any way to validate this theory? Also, checking the code I see that internally is doing storeEngine.get(key). Wouldn't it be more optimal to use getAll with blocks of keys? Regards, Carlos. ------=_Part_1482_23184810.1349210994555 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: 7bit Hi guys,

What's the expected performance of the AdminClient.fetchKeys? I'm using it but the behaviour is a bit heretic. I get something like 10K keys every 1.5 seconds, which is good, but then suddenly, the streaming stops for 30 to 60 seconds.

My suspicious is that the stops are caused by BDB reading, since some keys may not be cached. Is there any way to validate this theory? 

Also, checking the code I see that internally is doing storeEngine.get(key). Wouldn't it be more optimal to use getAll with blocks of keys?

Regards,
Carlos.
------=_Part_1482_23184810.1349210994555-- ------=_Part_1481_4044202.1349210994555--