[system test] Memcached segfault during rebalance with workload running causes rebalance failure

15G RAM 500G regular disc machine
cluster_ip: 172.23.105.23

Create a 8 node cluster, create default bucket with 6G RAM, 1 replica, sasl bucket with 5G RAM, 1 replica

Load items to both buckets into dgm state, around 70%.

Access the data for 2 hours, with workload:
default bucket: create:5,update:5,get:80,delete:5,expire:5,cache_miss:5, 10K ops/sec
saslbucket: create:5,update:5,get:20,delete:10,expire:60,cache_miss:5., 10K ops/sec

Then rebalance in one node with workload continue
Find core.memcached under /data path.
Saw a lot of errors like this:

Control connection to memcached on 'ns_1@172.23.105.23' disconnected: {{badmatch,
{error,
timeout}},
[

{mc_client_binary, cmd_binary_vocal_recv, 5}

,

{mc_client_binary, select_bucket, 2}

,

{ns_memcached, ensure_bucket, 2}

,

{ns_memcached, handle_info, 2}

,

{gen_server, handle_msg, 5}

,

{ns_memcached, init, 1}

,

{gen_server, init_it, 6}

,

{proc_lib, init_p_do_apply, 3}

]} (repeated 1 times)

Rebalance failed in a short time

This also happened in Windows the bug number is MB-8192
The stack trace of the core.memcached.xxxx on some node will be added later