PXC haproxy healthcheck, so many log in the pods, how to reduce it?

MySQL & MariaDB Percona XtraDB Cluster 8.x
,
1

Hi, I’ve managed to make pxc-haproxy working but I see 2 (or more) very long log string that display every second that I believe is the heath check, is this normal and is it affect the disk performance? How can I reduce the length of this or at least reduce the interval?


{"time":"05/Dec/2024:17:15:52.648", "client_ip": "10.244.0.3", "client_port":"39036", "backend_source_ip": "10.244.0.4", "backend_source_port": "37860", "frontend_name": "galera-in", "backend_name": "galera-nodes", "server_name":"baseapp-master-db-pxc-pxc-0", "tw": "1", "tc": "0", "Tt": "5", "bytes_read": "3509", "termination_state": "--", "actconn": "2", "feconn" :"2", "beconn": "1", "srv_conn": "1", "retries": "0", "srv_queue": "0", "backend_queue": "0" }
{"time":"05/Dec/2024:17:15:53.043", "backend_source_ip": "10.244.0.74", "backend_source_port": "33062", "message": "The following values are used for PXC node 10.244.0.74 in backend galera-nodes: wsrep_local_state is 4; pxc_maint_mod is DISABLED; wsrep_cluster_status is Primary; wsrep_reject_queries is NONE; wsrep_sst_donor_rejects_queries is OFF; 1 nodes are available"}
{"time":"05/Dec/2024:17:15:53.060", "backend_source_ip": "10.244.0.74", "backend_source_port": "33062", "message": "PXC node 10.244.0.74 for backend galera-nodes is ok"}

image
image1816×305 42 KB

2

just stumbled on the same issue, and looks like there’s an undocumented setting to suppress the health check logs

when the /etc/mysql/haproxy-env-secret/VERBOSE file has content “0”

If you are using a percona operator for mysql you can create a secret like

kind: Secret
apiVersion: v1
metadata:
  name: haproxy-secrets
  namespace: pxc
type: Opaque
stringData:
  VERBOSE: '0'

and modify deploy/cr.yaml by setting spec.haproxy.envVarsSecret: haproxy-secrets