Connection Limit Issue

Does the virtual service have a persistence method applied, and is the 'Override Connection Limit' option enabled for that persistence method? What type of virtual is configured?

Just to clarify, you're seeing the number of current sessions or max sessions exceeds the connection limit? By how much do they exceed the limit?

Hi Summers,

Any update?

I didn't have chance to apply the hotfix and do not know if this will solve. Unfortunately the problem still occurs. One of the pool members shows 107 affected connections. I see this value in the maximum number of connections.

The problem occurred again and I could see in time. When the problem first happened, I made some adjustments and now I'm using OnConnect mask 32. In the list of connections appeared many connections originating any6.any, and found SOL13850 to issue ID 284910. My VS is a standard so do not seem to make much sense what the SOL explains. I think I'll do the SOL workaround and see the result.

Any idea?

is there any special configuration such as irule?

can you post the configuration?

 tmsh list ltm virtual (name)
 tmsh list ltm pool (name)
 tmsh list ltm profile (type) (name)
 tmsh list ltm rule (name)

Hi nitass, there's nothing unusual in this VS. What I have realized is that always the last two members are beyond the limit.

    ltm virtual my_virtual_server {
    destination 1.1.1.26:http
    ip-protocol tcp
    mask 255.255.255.255
    pool my_pool
    profiles {
        http { }
        oneconnect_mask32 { }
        tcp { }
    }
    source 0.0.0.0/0
    source-address-translation {
        type automap
    }
    vlans {
        internal_vlan
    }
    vlans-enabled
    vs-index 36
}

ltm pool my_pool {
    load-balancing-mode least-connections-member
    members {
        1.1.1.254:ups-onlinet {
            address 1.1.1.254
            connection-limit 100
            session monitor-enabled
            state up
        }
        1.1.1.9:ups-onlinet {
            address 1.1.1.9
            connection-limit 100
            session monitor-enabled
            state up
        }
        1.1.1.217:ups-onlinet {
            address 1.1.1.217
            connection-limit 100
            session monitor-enabled
            state up
        }
        1.1.1.218:ups-onlinet {
            address 1.1.1.218
            connection-limit 100
            session monitor-enabled
            state up
        }
    }
    monitor http_monitor
    service-down-action reset
}

ltm profile one-connect oneconnect_mask32 {
    app-service none
    defaults-from oneconnect
    source-mask 255.255.255.255
}

there's nothing unusual in this VS.

yes, it looks straightforward configuration.

have you ever tried another load balancing algorithm?

Hello nitass, You may be right, I just saw that the two items that seem to always present problem, are shared in another pool with another balancing method. I'll change to "least connections node" and wait and see. Thx

Hi, can you tell me when have more connections from the server side than client side?