Sorry Willy I missed to add comments about all the helps from everyone, I should mention in the thread and correct my question more clear.
I searched about this issue and find some related thread. But most of them didn't provide detailed environment and I can't reproduce the environment to test and verify it. So I build a local test environment and provide the whole information to make sure other people can reproduce the issue. Another reason is that I find most thread just discuss modified some configure, then it just worked, but lack of the reason why it happened. I am interesting in the detailed logic about how the stick-table works. Since If I don't know the logic, I can't make sure my testing cases cover all the scenario. The mainly reason I created this thread is asking for the help about the internal logic of stick-table. 2017-02-22 14:42 GMT+08:00 Willy Tarreau <[email protected]>: > On Tue, Feb 21, 2017 at 09:29:42PM +0800, linbo liao wrote: > (...) > > If I change the table size to 100, and retest, in step5 all request are > > processed by p2. > > > > I don't know how stick-table to deal with concurrency connection, anyone > > can help to explain the stick-table working flow and why setting size to > 1 > > cause p1 accept some request? > > You asked me exactly all this privately and I responded to you, to which > you responded "Thank you for the clarification". If that was not clear, > it's better to say it than to ask the same question again to 1000 other > people. And if what I told you didn't work you can also say here "willy > told me this but after I tried it didn't work, has anyone any other > idea ?". People are participating on their own on their spare time, it's > important to respect their time by giving them all the context. Otherwise > the thread will start with the usual links to previous posts (the same > you forwarded me) and will waste everyone's valuable time. > > Willy >
