When I did the port of a Java AJP (Apache Jserv Protocol) from AIX to system Z Linux I used jmeter to conduct simulations for the sizing's.
We went from AJP http://en.wikipedia.org/wiki/Apache_JServ_Protocol To MOD_JK on system z Linux. http://tomcat.apache.org/tomcat-3.3-doc/mod_jk-howto.html Case and point, being when you conduct research for porting you always bring your source technology stack forward to relatively recent levels during the migration. It should never (IMHO) be a one for one level migration. We have also ported *many* other workloads from XXX to LoZ - Some approaches: I would 1) Port the stacks with current OS, WEB,APP,DB servers. 2) Study the zPlatform specific well documented OS settings * z/VM subsystems (Processor, storage, I/O, paging) * z/VM guest resource * Linux on System z - disks, storage, processor * Networked servers (Linux, Unix, Windows) * Linux Applications 3) Build a driver tool that allows you to meet the desired #Users and concurrency. 4) Execute and acquire. During the port you may wish to over configure (CPU_MEM) and then scale back as you iterate the tests. http://jakarta.apache.org/jmeter/ http://wiki.apache.org/jakarta-jmeter/JMeterLinks Gerard C. Shockley Boston University [email protected] 617.353.9898 (w) 617.353.6171 (f) http://www.americaspower.org/The-Facts -----Original Message----- From: Linux on 390 Port [mailto:[email protected]] On Behalf Of Jim Moling Sent: Friday, March 13, 2009 12:40 PM To: [email protected] Subject: Server sizing from AIX to z/Linux - Rules of thumb? Wondering if anyone has some rules-of-thumb for sizing a z/Linux virtual server for AIX-based workloads, specifically CPU & memory. All I am looking for are rough guidelines that can be used as a starting point to estimate resources for moving a workload from a mid-tier platform (AIX) to a mainframe virtual Linux server. For example, we have AIX Server based applications that use a 3-tier architecture that consists of: 1) Web Server - 2 CPU, 2 GB RAM 2) App Server - 4 CPU, 8 GB RAM 3) DB Server - 4 CPU's, 16 GB RAM, 100 GB Storage I realize that it ultimately depends on just how busy a server is in order to determine resource requirements with any degree of accuracy, and explained as much to those that asked me about this, and being short of any empirical data at this point, was hoping that others may have derived some basic guidelines based on their experience. We're looking at potentially moving hundreds of AIX servers onto mainframe-based Linux servers, so trying to get a handle on just what that will mean in terms of resources. I figure the disk-based requirements would be pretty much Gig-for-Gig, but not sure how to factor AIX CPU's to z10 IFL's and the scaling factors going from AIX RAM to central storage. I humbly ask for any help towards getting a better handle on this, which would be greatly appreciated. Thanks, Jim Moling ----------------------------------------- This E-mail and its attachments (if any) are intended solely for the use of the addressee(s) and may contain sensitive but unclassified information. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or use of the information contained herein (including any reliance thereon) is strictly prohibited. If you have received this E-mail in error, please notify the sender immediately and destroy the E-mail and any attachments. ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
