Yeah I will do. On OSX "top" doesn\t show a "wait io", so I need to find out how to get this.
I will also get ntop installed to monitor the traffic. What I can say is that the machine has 4G, and is happily staying under the 4G usage and CPU is around 50%. It is a cheap giga switch and I get 30M/sec sustained transfer between the nodes when scp'ing, so I presume it to be waiting on disk IO I'll try and profile though tomorrow Cheers Tim On Thu, Jul 23, 2009 at 8:01 PM, stack<[email protected]> wrote: > On Thu, Jul 23, 2009 at 7:08 AM, tim robertson > <[email protected]>wrote: > >> >> So trying to isolate the MapReduce read, I comment out the >> //table.put(row); in the Map and have the same demons running on the >> cluster (e.g. HBase sitting idle), it reads at 18,500/sec >> >> Would it be correct to assume this is just a limitation of the disks >> on this cluster since writing alone seems reasonable and reading alone >> seems reasonable? >> > > Tim, can you profile the hardware when reading and writing? Whats it > doing? Are clients just waiting on i/o? > St.Ack >
