During your tests, what is your cache hit ratio? GWC should with some fairly predictable tweaks (number of threads supported by Tomcat, and so forth) be able to handle around 4000 tiles per second, at which point gigabit ethernet will be saturated, for most tiles.
-Arne On 10/1/10 1:13 PM, DGIS Devels wrote: > Hi, we have a question about Geoserver performance. > > We have been trying for a couple of months to find a good > configuration that offers a solution to our use case. You can see our > current architecture in the attachment: > - WebLogic cluster with 2 machines and 4 managed servers (2 per > machine). We have tested it whit JRockit and Sun JVM. > - 4 Geoservers (one per managed server) > - 2 GeoWebCaches (one per machine) > - A software balancer which balances requests between GeoWebCaches and > Geoservers > > We've testing too with a lot of values c for parameters like memory, > render memory, timeouts, DB connections ... > > Our goal is to support 85 concurrent users performing an average of 5 > map requests (near of 100 256x256 tiles) in a time of 10 minutes. It > must be equivalent, according to our calculations, to respond to some > 85000 requests in those 10 minutes. > > So far, the max number of request we've been able to response is about > 30000 tiles. In most cases, some GeoServers or GeoWebCaches instances > crashes. > > Anybody has made similar tests? Do you think is possible to reach a > similar performance? > ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing. http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ Geoserver-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/geoserver-users
