RE: ActiveMQ on Physical Server vs. VM

2014-01-28 Thread barry.barnett
g Subject: Re: ActiveMQ on Physical Server vs. VM Like with pretty much any Java app, you end up running a VM in a VM. Performance on IO, CPU sharing and so on will be impacted. That doesn't mean that you couldn't scale in different ways with more VM instances instead. On Jan 24, 2014

Re: ActiveMQ on Physical Server vs. VM

2014-01-24 Thread Johan Edstrom
Like with pretty much any Java app, you end up running a VM in a VM. Performance on IO, CPU sharing and so on will be impacted. That doesn't mean that you couldn't scale in different ways with more VM instances instead. On Jan 24, 2014, at 6:37 AM, artnaseef wrote: > Interesting question (

RE: ActiveMQ on Physical Server vs. VM

2014-01-24 Thread artnaseef
Interesting question (dedicatd NIC) - that's more a VM question than an ActiveMQ question. With sharing the NIC with other load, the issue becomes, what other loads are sharing and how much? Very much a question outside of ActiveMQ itself. Is it possible for a VM host to dedicate a physical NIC

RE: ActiveMQ on Physical Server vs. VM

2014-01-24 Thread barry.barnett
Message- From: artnaseef [mailto:a...@artnaseef.com] Sent: Thursday, January 23, 2014 1:53 PM To: users@activemq.apache.org Subject: Re: ActiveMQ on Physical Server vs. VM No more advantage than for any other application/service. There's nothing inherit to ActiveMQ that needs real har

Re: ActiveMQ on Physical Server vs. VM

2014-01-23 Thread artnaseef
No more advantage than for any other application/service. There's nothing inherit to ActiveMQ that needs real hardware - it relies entirely on the O/S to handle low-level hardware I/O. -- View this message in context: http://activemq.2283324.n4.nabble.com/ActiveMQ-on-Physical-Server-vs-VM-tp46