we use varnish 3.02 with cache server, and have 6 and more servers as backends with different weight ,so round-robin may not for us,
we use random director with below health check, but it seems not working
probe apachehealthcheck {
.url = "/test/test.php";
.interval = 3s;
.timeout = 3 s;
.window = 5;
.threshold = 3;
.initial = 3;
.expected_response = 200; }
backend passport backend passport1 {
.host = "192.168.2.192";
.port = "80";
.connect_timeout = 10s;
.first_byte_timeout = 50s;
.between_bytes_timeout = 30s;
.probe = apachehealthcheck; }
director passport_cluster random {
{.backend = flpassport1;.weight = 1;}
{.backend = flpassport2;.weight = 2;}
... }
i saw the lines below at https://www.varnish-cache.org/trac/wiki/BackendPolling
The random director will not consider backends which are unhealthy part of the pool.
does it mean when i am useing random director in varnish, when some backends broken, it still forward the request to ?