Squid health monitoring via Telegraf Influxdb Grafana

Example Grafana dashboards

Idea

Squid shares its statistics via SNMP. Metrics are scraped by another host using Telegraf SNMP feature. Telegraf saves metrics into InfluxDB. Charts are presented in Grafana.

Basic setup

1. Host with installed squid shares stats via built-in SNMP. Configuration:

   a.) add content squid_host/etc/squid/squid.conf to your /etc/squid/squid.conf

   b.) Open 3401 udp port from IP Address where Telegraf runs

   c.) Modify line "acl snmpnet src YOUR_TELEGRAF_IP_ADDRESS" inside /etc/squid/squid.conf replacing string YOUR_TELEGRAF_IP_ADDRESS with valid hostname/IP.

   d.) Restart Squid service e.g. ```bash systemctl restart squid ```

2. Telegraf host:

   a.) Place telegraf_influxdb_host/etc/telegraf/telegraf.conf to /etc/telegraf/ (or add only section [[inputs.snmp]] )

   b.) Adjust line "agents = [ YOUR_SQUID_IP_ADDRESS:3401" ]" by setting valid hostname or IP of Squid address.

   c.) Restart telegraf service e.g. ```bash systemctl restart telegraf ```

3. Grafana

   1. Ensure InfluxD database named "telegraf" is set as a Data Source in Grafana Configuration

   2. Import Squid_Dashboard_01.json into Grafana

TODO

• add memory usage of all Squid processes to graphs