Exportd and DocAPI - Dev Survey

Community Support
1

Hi everyone,
we are planning to implement an advanced reporting system in DATAGerry which will most probably significantly affect the current Exportd and DocAPI features.

Therefore we would like to know about your experiences and usages of the mentioned features with some questions.

  1. Do you use Exportd / DocAPI features at all?
  2. If yes, what are you using Exportd/DocAPI for (some bullet points would be nice) ?
  3. Which external systems are you communicating with when using Exportd (like OpenNMS, Anisble and the others available )?
  4. Which features are still required from your point of view regarding a reporting system ?

BR Adnan

2

Hi Adnan,

thanks for asking. While we don’t use the DocAPI, Exportd is very important in our environment. We use the following external systems:

  • ExternalSystemAnsible - for connecting our Ansible setup
  • ExternalSystemGenericPullJson - for connecting our Prometheus monitoring

With Datagerry, Ansible and Prometheus we manage the whole workflows for our Linux and Windows servers. At the moment, these workflows are:

  • Deploying/Installing a new Windows/Linux server
  • Installing security patches/updates on all of the servers
  • Resizing a server (extending RAM/CPU cores automatically at night)

Installing a new server:

image
image819×919 42.4 KB

When installing a new server, we create at first an object in Datagerry and set the field “Status” (state) to “provision” and fill all the details like IP, hostname, OS, VLAN, … After that, we start an Ansible playbook, which creates a new virtual machine, installs the operating system and required base software. When the setup is done, Ansible sends a REST calll to the Datagerry API and sets the state of the server to “productive”. Also the monitoring field will be set from “nein” (no) to “ja” (yes). This will cause that the server is part of the export to our Prometheus setup.

Installing security patches/updates:

For installing security updates on our systems, we organize our servers in patch groups (“Patchwellen”), which we handle with a field in Datagerry. With Ansible we get all servers of a patchgroup and install security updates at night.

Resizing a server:
Sometimes we need to resize a server, to extend memory or CPU cores. We set the field “Aktion” (action) to “Resize VM” in Datagerry and configure the fields cpu cores and RAM with the new values. At night, Ansible we check all servers with the “Resize VM” field, shut them down, change the VM configuration, and bring them back online

3

Hi @michaelb,
Thanks for the feedback!

BR Adnan

4

Hello Adnan,
We do not use docAPI, and we work in a very similar way as @michaelb: an Ansible playbook is used to build our Windows and Linux servers, and it uses API to document our infra in the CMDB.
We also group servers into patch groups, for security updates.

For our reporting, so far we use API to build reports like “list servers without IP”, or “list application services without servers” or to extract CMDB data with reference data.

Kind regards.