Typically you want 2-3 partitions per CPU core to get good load balancing. How big is the data you’re transferring in this case? And have you looked at the machines to see whether they’re spending lots of time on IO, CPU, etc? (Use top or dstat on each machine for this). For large datasets with larger numbers of tasks, one option we added in 0.8.1 that helps a lot is consolidating shuffle files (see http://spark.incubator.apache.org/releases/spark-release-0-8-1.html). However, another common problem is just serialization taking a lot of time, which you’ll notice if the application is CPU-heavy, and which you can fix using Kryo.
Matei On Jan 9, 2014, at 2:11 PM, Yann Luppo <[email protected]> wrote: > Thank you guys that was really helpful in identifying the slow step, which in > our case is the leftouterjoin. > I'm checking with our admins to see if we have some sort of distributed > system monitoring in place, which I'm sure we do. > > Now just out of curiosity, what would be the rule of thumb or general > guideline for the number of partitions and the number of reducers? > Should it be some kind of factor of the number of cores available? Of nodes > available? Should the number of partitions match the number of reducers or at > least be some multiple of it for better performance? > > Thanks, > Yann > > From: Evan Sparks <[email protected]> > Reply-To: "[email protected]" <[email protected]> > Date: Wednesday, January 8, 2014 5:28 PM > To: "[email protected]" <[email protected]> > Cc: "[email protected]" <[email protected]> > Subject: Re: performance > > On this note - the ganglia web front end that runs on the master (assuming > you're launching with the ec2 scripts) is great for this. > > Also, a common technique for diagnosing "which step is slow" is to run a > '.cache' and a '.count' on the RDD after each step. This forces the RDD to be > materialized, which subverts the lazy evaluation that causes such diagnosis > to be hard sometimes. > > - Evan > > On Jan 8, 2014, at 2:57 PM, Andrew Ash <[email protected]> wrote: > >> My first thought on hearing that you're calling collect is that taking all >> the data back to the driver is intensive on the network. Try checking the >> basic systems stuff on the machines to get a sense of what's being heavily >> used: >> >> disk IO >> CPU >> network >> >> Any kind of distributed system monitoring framework should be able to handle >> these sorts of things. >> >> Cheers! >> Andrew >> >> >> On Wed, Jan 8, 2014 at 1:49 PM, Yann Luppo <[email protected]> wrote: >> Hi, >> >> I have what I hope is a simple question. What's a typical approach to >> diagnostic performance issues on a Spark cluster? >> We've followed all the pertinent parts of the following document already: >> http://spark.incubator.apache.org/docs/latest/tuning.html >> But we seem to still have issues. More specifically we have a leftouterjoin >> followed by a flatmap and then a collect running a bit long. >> >> How would I go about determining the bottleneck operation(s) ? >> Is our leftouterjoin taking a long time? >> Is the function we send to the flatmap not optimized? >> >> Thanks, >> Yann >>
