Error allocating space for multicast recv buffer

Hi,

We have the following cluster configuration :-

Number of instances : 3

**Namespace Configuration : **

namespace app-persistent {

        replication-factor 2
        memory-size 3G
        default-ttl 30d # 30 days, use 0 to never expire/evict.
        ldt-enabled true
        storage-engine device {
           file /data/persistent.dat
           filesize 35G
        }
}

namespace app-nonpersistent {

        replication-factor 2
        memory-size 5G
        default-ttl 30d # 30 days, use 0 to never expire/evict.
        storage-engine memory
}

**Estimated hits / second : 300 **

We have a mesh cluster configuration.

As soon as released we our application, in a few mins, the application started throwing ‘seed node down’, and in aerospike we saw the following error :-

Aug 10 2017 12:48:56 GMT: WARNING (hb): (hb.c:4040) error allocating space for multicast recv buffer of size 33619974 on fd 122

Can anyone please help here ??

Could you share the following details.

  1. aerosike-server version
  2. has aerospike service crashed on any of the machines.
  3. is port 3002 on these machines publicly visible.
  4. If possible the entire configuration file. The logs seem to indicate that you are running multicast instead of mesh.

From the log message it looks like there is some non-aerospike traffic hitting the heartbeat port 3002 on these machines.