Reporting with Grafana

INFO

This guide is about Grafana. If you want to use another tool, check if it supports Elasticsearch indexes. If yes, then you are good to go!

Grafana is a third-party analytics/visualization application you can connect to Zammad (precisely: Elasticsearch). It can access the Elasticsearch index and visualize your Zammad data.

This guide will provide you some steps to get started. For a deeper insight you should consider having a look at Elasticsearch's indexed attributes and to read the documentation of Grafana.

This guide expects all requirements to be up and running. We will not cover core configurations of each tool. Please also note that we can't support you with configuration of your specific third party tool.

Prerequisites

You need:

• A running instance of Grafana (hosted or self-hosted) in version 10.3 or higher
• Read access to your Elasticsearch index
• A Zammad instance in version 4 or higher

WARNING

Never expose Elasticsearch to the public if you're not sure how to do it. Especially never without authentication! Zammad stores very sensitive information within the Elasticsearch Index.

Setting up Required Data Sources

Before we start: The data sources always follow the same scheme. We reduced below information to name, time field name and index name. Everything else relies on your environment and is out of our scope.

TIP

Replace zammad_production_ with your fitting prefix.

How to query the index?

Adjust the following command to your environment:

curl http://localhost:9200/_aliases?pretty=true

This will return an output like the following:

{
  "zammad_production_knowledge_base_translation" : {
    "aliases" : { }
  },
  "zammad_production_ticket_priority" : {
    "aliases" : { }
  },
  "zammad_production_stats_store" : {
    "aliases" : { }
  },
  "zammad_production_organization" : {
    "aliases" : { }
  },
  "zammad_production_cti_log" : {
    "aliases" : { }
  },
  "zammad_production_group" : {
    "aliases" : { }
  },
  "zammad_production_knowledge_base_answer_translation" : {
    "aliases" : { }
  },
  "zammad_production_ticket" : {
    "aliases" : { }
  },
  "zammad_production_ticket_state" : {
    "aliases" : { }
  },
  "zammad_production_chat_session" : {
    "aliases" : { }
  },
  "zammad_production_user" : {
    "aliases" : { }
  },
  "zammad_production_knowledge_base_category_translation" : {
    "aliases" : { }
  }
}

ES - Chat Sessions

• Index name: zammad_production_chat_session
• Time field name: created_at

ES - CTI Log

• Index name: zammad_production_cti_log
• Time field name: start_at

ES - Ticket Articles

• Index name: zammad_production_ticket
• Time field name: article.created_at

ES - Tickets by closed_at

• Index name: zammad_production_ticket
• Time field name: close_at

ES - Tickets by created_at

• Index name: zammad_production_ticket
• Time field name: created_at

ES - Tickets by first_response_at

• Index name: zammad_production_ticket
• Time field name: first_response_at

With above data sources you basically have everything you need to start building your own dashboards.

Quick Start with Dashboard Template

If you want to get inspired, you can use our sample dashboards as mentioned below. These dashboards can also be found on GitHub.

Importing a Dashboard

In Grafana, select ➕ > Import (or any other place which offers you to import a dashboard) and either upload the json file you downloaded from Github or use the grafana.com ID, provided as badge like 12345 attached to the next sections titles.

During importing you can provide a dashboard name and folder. You'll also be asked to map the data sources to your environment. If you used our data source names above, you can simply search for the same name.

Ticket Statistics Dashboard 14222

Dashboard Graphs

• ticket opening and closing^[1]
• created articles
• ticket SLA (in time and violation) per type^[1:1][2]

Ticket and Article Meta Information

• ticket group distribution
• sender ratio (e.g. Customer / Agent)^[3]
• article type ratio (e.g. email, phone)^[3:1]
• article content type
• escalation ratios^[1:2]
• average first response, update time and close time^[2:1]
• top 10 of:
  • organization of ticket customer^[1:3]
  • ticket customers^[1:4]
  • ticket owners^[1:5]
  • average accounted time on ticket
  • ticket tags^[1:6]
• last 10 escalated tickets

Required Data Sources

• ES - Ticket Articles
• ES - Tickets by created_at
• ES - Tickets by closed_at

Chat-Session Statistics Dashboard 14224

Dashboard Graphs

Chat session creations.

Chat Session Meta Information

• top 10 of:
  • chat tags
  • chat agents
  • chat exit pages
  • city origins
• chat topic ratio
• average number of messages within chat-sessions
• average chatting time
• world map with chat origin countries

Required Data Sources

• ES - Chat Sessions

CTI-Log Statistics Dashboard 14223

Dashboard Graphs

Number of calls per direction (in / out).

Chat Session Meta Information

• call ratio (in / out)
• average waiting time
• average talking time
• top 10 of:
  • callers (in)
  • call answerers (in)

Required Data Sources

• ES - CTI Log

---

1. Some values are not available as time series information. This means we can only display the last value of the field in question. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

2. Requires SLA function to be active. Negative values indicate SLA violations. ↩︎ ↩︎

3. Specific reference IDs are not the same on every instance and thus the panel may not work or show incorrect data. Check the panels description on how to find our the relations on your system. ↩︎ ↩︎