Lot of error entries "Lost connection to MySQL server during query"

[cybermcm]

Searched forum and google for this and found entries but no real solution.
Zabbix is running and working, but there are a lot of error messages like

[Z3005] query failed: [2013] Lost connection to MySQL server during query

This happens between 1 and 10 minutes.

My setup:
Raspi 4 with Raspi OS (all updates) and docker

sql:
  container_name: sql
  environment:
   #- PUID=1000 # mariadb
   #- PGID=1000 # mariadb
   - MYSQL_ROOT_PASSWORD=xxx
   - TZ=xxx # mariadb
  hostname: sql
  image: ghcr.io/linuxserver/mariadb
  labels:
   - com.centurylinklabs.watchtower.enable=true
  networks:
   - compose
  ports:
   - "3306:3306"
  restart: always
  volumes:
   - /etc/timezone:/etc/timezone:ro
   - /etc/localtime:/etc/localtime:ro
   - /_sql:/config # mariadb
   - /opt/docker/_backup:/opt/docker/_backup

zabbix:
   container_name: zabbix
   depends_on:
    - sql
    - zabbix-server
   environment:
    - PHP_TZ=xxx
    - DB_SERVER_HOST=sql
    - MYSQL_ROOT_PASSWORD=xxx
   hostname: zabbix
   image: zabbix/zabbix-web-nginx-mysql:alpine-latest
   labels:
    - com.centurylinklabs.watchtower.enable=true  
   networks:
    - compose
   ports:
    - "8083:8080"
   restart: always
   volumes:
   - /etc/timezone:/etc/timezone:ro
   - /etc/localtime:/etc/localtime:ro
 
zabbix-agent:
  container_name: zabbix-agent
  depends_on:
   - zabbix-server
  environment:
   - ZBX_HOSTNAME=xxx
   - ZBX_SERVER_HOST=172.19.1.2
   #- ZBX_ACTIVESERVERS=zabbix-server
   #- ZBX_DEBUGLEVEL=4
  hostname: zabbix-agent
  #image: zabbixmultiarch/zabbix-agent2
  image: zabbix/zabbix-agent2:alpine-latest
  labels:
   - com.centurylinklabs.watchtower.enable=true  
  network_mode: host
  #networks:
   #- compose
  ports:
   - "10050:10050"
  privileged: true
  restart: always
  user: 0:0
  volumes:
   - /etc/timezone:/etc/timezone:ro
   - /etc/localtime:/etc/localtime:ro
   - /var/run/docker.sock:/var/run/docker.sock:ro
 
zabbix-server:
  container_name: zabbix-server
  depends_on:
   - sql
  environment:
   - PHP_TZ=xxx
   - DB_SERVER_HOST=sql
   - MYSQL_ROOT_PASSWORD=xxx
   - MYSQL_USER=zabbix
   - MYSQL_PASSWORD=xxx
   - MYSQL_DATABASE=zabbix
   - ZBX_STARTVMWARECOLLECTORS=5
  hostname: zabbix-server
  image: zabbix/zabbix-server-mysql:alpine-latest
  labels:
   - com.centurylinklabs.watchtower.enable=true  
  networks:
   compose:
    ipv4_address: 172.19.1.2
  ports:
   - "10051:10051"
  restart: always
  volumes:
   - /etc/timezone:/etc/timezone:ro
   - /etc/localtime:/etc/localtime:ro

With the armv7 platform there is nor real stable mysql container, used an older mysql container -> no errors; switching to current MariaDB container (for ARMv7, 10.1.48-MariaDB-0ubuntu0.18.04.1 - Ubuntu 18.04) the error is there

Tried to lower keepalive, but this didn't solve the issue

net.ipv4.tcp_keepalive_time=300

any ideas (again, platform is working, trying to clear unnecessary errors)?

[jeschebach]

I'm having the same error and I have a zabbix version 4.4.8, running over alpine linux and connected to an RDS Aurora DB.
Over 450 hits during the last 24hr.

The ping test does no reveal any networking issue at first glance:

--- xxx ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 1017097ms
rtt min/avg/max/mdev = 0.704/0.819/26.769/0.878 ms
​

The DB has the following timeout parameters:

+-------------------------------------------+----------+
| Variable_name                             | Value    |
+-------------------------------------------+----------+
| aurora_fwd_master_idle_timeout            | 240      |
| aurora_globaldb_rpo_wait_timeout          | 60       |
| aurora_zdr_timeout_on_replica_fall_behind | 60       |
| connect_timeout                           | 540      |
| delayed_insert_timeout                    | 300      |
| have_statement_timeout                    | YES      |
| innodb_flush_log_at_timeout               | 1        |
| innodb_lock_wait_timeout                  | 50       |
| innodb_rollback_on_timeout                | OFF      |
| interactive_timeout                       | 360      |
| lock_wait_timeout                         | 31536000 |
| net_read_timeout                          | 720      |
| net_write_timeout                         | 300      |
| rpl_stop_slave_timeout                    | 31536000 |
| slave_net_timeout                         | 60       |
| wait_timeout                              | 86400    |
+-------------------------------------------+----------+​

Any ideas or suggestions to look for?