[Warning] Aborted connection

Hello,

Follow our my.cnf and error that occurs in 3 clusters of replication that have created with Percona.

Is only generated the error, we have no packet losses, because we made a ping and even called me back normally.

Have increased the max_allowed_packet and not resolved.

ERRO:

140211 9:06:24 [Warning] Aborted connection 4229297 to db: ‘DB’ user: ‘USER’ host: ‘10.1.1.1’ (Got an error reading communication packets)
140211 9:06:46 [Warning] Aborted connection 4229572 to db: ‘DB’ user: ‘USER’ host: ‘10.1.1.1’ (Got an error reading communication packets)
140211 9:07:43 [Warning] Aborted connection 4230285 to db: ‘DB’ user: ‘USER’ host: ‘10.1.1.1’ (Got an error reading communication packets)

MY.CNF

[mysqld]

server_id=1

binlog_format=ROW

log_bin=mysql-bin

wsrep_cluster_address=gcomm://NODE1,NODE2,NODE3

wsrep_provider=/usr/lib64/libgalera_smm.so

datadir=/var/lib/mysql

wsrep_cluster_name=NAMECLUSTER

wsrep_sst_method=xtrabackup-v2

wsrep_node_name=NODE1

log_slave_updates

innodb_locks_unsafe_for_binlog=1

innodb_autoinc_lock_mode=2

wsrep_sst_auth=“sstuser:sst12345”

#Variaveis

wsrep_provider_options=“gcache.size=2G;”
wsrep_slave_threads=16
wsrep_causal_reads=ON
#wsrep_replicate_myisam=1

The MySQL server

[mysqld]
port = 31890
#socket = /var/lib/mysql/mysql.sock
key_buffer = 5M
max_allowed_packet = 512M
table_cache = 64
sort_buffer_size = 512K
net_buffer_length = 8M
read_buffer_size = 256K
read_rnd_buffer_size = 512K
max_tmp_tables = 100
join_buffer_size = 1M
max_join_size = 1M
max_connections = 500
max_connect_errors = 500
max_join_size = 10M
log_warnings = 2
long_query_time = 20
open_files_limit = 65000
myisam_sort_buffer_size = 3M
wait_timeout = 6000
#read_timeout = 600
connect_timeout = 6000
query_cache_type = 0
query_cache_size = 0

interactive_timeout = 6000

net_read_timeout = 6000
net_write_timeout = 6000

expire_logs_days = 7

Regras customizadas no my.cnf

#tmp_table_size = 16M
#max_heap_table_size = 16M
#thread_cache_size = 8
#table_cache = 64
#max_connections = 600

join_buffer_size = 1M
tmp_table_size = 16M
max_heap_table_size = 16M
key_buffer_size = 16M
#innodb_force_recovery = 1

#plugin-load=thread_pool.so
thread_pool_size = 24

#skip-networking
bind-address=IP-NODE
skip-name-resolve

[InnoDB]
innodb_flush_method = O_DIRECT
innodb_read_io_threads = 16
innodb_write_io_threads = 16
innodb_additional_mem_pool_size = 64M
innodb_log_buffer_size = 512M
#innodb_data_file_path=ibdata1:1GB:autoextend

[safe_mysqld]
#err-log=/var/lib/mysql/mysqld.log
wsrep_url=gcomm://NODE1,gcmm://NODE2,gcomm://NODE3,gcomm://

[mysqldump]
quick
max_allowed_packet = 64M

[mysql]
no-auto-rehash

[isamchk]
key_buffer = 7M
sort_buffer_size = 7M
read_buffer = 2M
write_buffer = 2M

[myisamchk]
key_buffer = 7M
sort_buffer_size = 7M
read_buffer = 2M
write_buffer = 2M

[mysqlhotcopy]
#interactive-timeout

Do you maybe have large number of connections per second? Try this article if it helps:
[url]http://www.mysqlperformanceblog.com/2012/01/06/mysql-high-number-connections-per-secon/[/url]

Or any firewall/selinux issue…? do a telnet that node for 4567 port and see if u r able to connect.?