thor533 commented on issue #9225:
URL: https://github.com/apache/apisix/issues/9225#issuecomment-1510660668
> > > The current implementation may suffer from imbalance at low quantity
levels. Currently using the heap to maintain the connection is only balanced
within a single worker. Can you try using a single Nginx worker?
> >
> >
> > Is it normal for you to test?
>
> I write a unit test and run it on a local machine. **it's working
normally**.
>
> the unit test is as follows. ( actually, I modify the original test, which
files you find from `t/node/least_conn.t`)
>
> ```lua
> === TEST 2: select least conn
> --- apisix_yaml
> upstreams:
> -
> id: 1
> type: least_conn
> nodes:
> "127.0.0.1:1980": 1
> "127.0.0.2:1980": 1
> "127.0.0.3:1980": 1
> "127.0.0.4:1980": 1
> --- config
> location /t {
> content_by_lua_block {
> local http = require "resty.http"
> local uri = "http://127.0.0.1:"; .. ngx.var.server_port
> .. "/mysleep?seconds=0.1"
>
> local t = {}
> for i = 1, 3000 do
> local th = assert(ngx.thread.spawn(function(i)
> local httpc = http.new()
> local res, err = httpc:request_uri(uri..i, {method =
"GET"})
> if not res then
> ngx.log(ngx.ERR, err)
> return
> end
> end, i))
> table.insert(t, th)
> end
> for i, th in ipairs(t) do
> ngx.thread.wait(th)
> end
> }
> }
> --- request
> GET /t
> --- grep_error_log eval
> qr/proxy request to \S+ while connecting to upstream/
> --- grep_error_log_out
> proxy request to 127.0.0.1:1980 while connecting to upstream
> proxy request to 0.0.0.0:1980 while connecting to upstream
> proxy request to 127.0.0.1:1980 while connecting to upstream
> ```
>
> try to run it, and ignore the error output.
>
> ```shell
> $ prove -I. -I../test-nginx/inc -I../test-nginx/lib -r t/node/least_conn.t
> ```
>
> Use the log to count the back-end distribution
>
> ```shell
> $ cat t/servroot/logs/error.log |grep 'while connecting to upstream' |
grep -o 'run():.*client'|grep -o 'run(): proxy request to 127\.0\.0\.[1-3]'|
sort | uniq -c
> 773 run(): proxy request to 127.0.0.1
> 770 run(): proxy request to 127.0.0.2
> 843 run(): proxy request to 127.0.0.3
> ```
However, when I tested the least-conn algorithm in a centos environment, I
got exactly the same results as the polling algorithm,
1.First server-->10.xx.xx.12,
the conn number is 92,
2.Second server -->10.xx.xx.105
the conn number is 3
But the result is the same as the weighted polling algorithm
Is there something wrong with me
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: notifications-unsubscr...@apisix.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
