Re: [infinispan-dev] Store as binary

2014-02-05 Thread Mircea Markus
On Feb 4, 2014, at 7:14 AM, Galder Zamarreño gal...@redhat.com wrote: On 21 Jan 2014, at 17:45, Mircea Markus mmar...@redhat.com wrote: On Jan 21, 2014, at 2:13 PM, Sanne Grinovero sa...@infinispan.org wrote: On 21 January 2014 13:37, Mircea Markus mmar...@redhat.com wrote: On Jan

Re: [infinispan-dev] Store as binary

2014-01-29 Thread Radim Vansa
20 % writes, 80 % reads Radim On 01/29/2014 03:20 PM, Paul Ferraro wrote: What was the read/write ratio used for this test? On Fri, 2014-01-17 at 14:06 +0100, Radim Vansa wrote: Hi Mircea, I've ran a simple stress test [1] in dist mode with store as binary (not enabled, enabled keys only,

Re: [infinispan-dev] Store as binary

2014-01-21 Thread Galder Zamarreño
On Jan 21, 2014, at 1:36 PM, Sanne Grinovero sa...@infinispan.org wrote: What's the point for these tests? +1 On 20 Jan 2014 15:48, Radim Vansa rva...@redhat.com wrote: OK, I have results for dist-udp-no-tx or local-no-tx modes on 8 nodes (in local mode the nodes don't communicate,

Re: [infinispan-dev] Store as binary

2014-01-21 Thread Sanne Grinovero
On 21 January 2014 13:37, Mircea Markus mmar...@redhat.com wrote: On Jan 21, 2014, at 1:21 PM, Galder Zamarreño gal...@redhat.com wrote: What's the point for these tests? +1 To validate if storing the data in binary format yields better performance than store is as a POJO. That will

Re: [infinispan-dev] Store as binary

2014-01-21 Thread Mircea Markus
On Jan 21, 2014, at 2:13 PM, Sanne Grinovero sa...@infinispan.org wrote: On 21 January 2014 13:37, Mircea Markus mmar...@redhat.com wrote: On Jan 21, 2014, at 1:21 PM, Galder Zamarreño gal...@redhat.com wrote: What's the point for these tests? +1 To validate if storing the data in

Re: [infinispan-dev] Store as binary

2014-01-20 Thread Mircea Markus
Hi Radim, I think 4 nodes with numOwner=2 is too small of a cluster. My calculus here[1] points out that for numOwners=1, the performance benefits is only visible for clusters having more than two nodes. Following a similar logic for numOwenrs=2, the benefit would only be visible for clusters

Re: [infinispan-dev] Store as binary

2014-01-20 Thread Pedro Ruivo
Hi, IMO, we should try the worst scenario: Local Mode + Single thread. this will show us the highest impact in performance. Cheers, Pedro On 01/20/2014 09:41 AM, Mircea Markus wrote: Hi Radim, I think 4 nodes with numOwner=2 is too small of a cluster. My calculus here[1] points out that

Re: [infinispan-dev] Store as binary

2014-01-20 Thread Mircea Markus
Would be interesting to see as well, though performance figure would not include the network latency, hence it would not tell much about the benefit of using this on a real life system. On Jan 20, 2014, at 9:48 AM, Pedro Ruivo pe...@infinispan.org wrote: IMO, we should try the worst scenario:

Re: [infinispan-dev] Store as binary

2014-01-20 Thread Radim Vansa
OK, I have results for dist-udp-no-tx or local-no-tx modes on 8 nodes (in local mode the nodes don't communicate, naturally): Dist mode: 3 % down for reads, 1 % for writes Local mode: 19 % down for reads, 16 % for writes Details in [1], ^ is for both keys and values stored as binary. Radim [1]

[infinispan-dev] Store as binary

2014-01-17 Thread Radim Vansa
Hi Mircea, I've ran a simple stress test [1] in dist mode with store as binary (not enabled, enabled keys only, enabled values only, enabled both). The difference is 2 % (with storeAsBinary enabled fully being slower). Radim [1]