Home > Enterprise >  The was really better than single clustering performance?
The was really better than single clustering performance?

Time:09-18

Two machines configuration are 8 c, 16 g, swap16g, the operating system is a Red Hat Enterprise Linux 6.8, was version is V8.5.5.11, create a cluster, a total of six cluster nodes, F5 to do load balancing, of which the DMGR built on one machine, a machine three nodes, the same application functions, in I run on the local computer takes 4 seconds, in the test environment is one of the node takes 6 seconds no F5 (test environment), my computer configuration is i5 processor, 8 gb of memory, why the testing environment and my own local gap is so big, beg god to explain!
Node configuration:
1, to enable the servlet caching
2, the data source using ali Druid
Threads 3, minimum, maximum, respectively is 40 reached.
4, the initial Java virtual machine heap minimum, maximum, respectively is 1024 m, 3072 m
All node configuration,

CodePudding user response:

come on

CodePudding user response:

CodePudding user response:

  • Related