QA

Motivation¶

During the SUSE DCT migration 2022-2024 #121720 new networks along with a new setup for DHCP/DNS/PXE was implemented. Some existing legacy resources persisted not being fully aligned with the new setup, e.g. fozziebear as DHCP and potentially also DNS server for qe.prg2.suse.org was migrated from the former location. Also qanet.qa.suse.cz in PRG1 remained untouched. To address the painpoints of differing setup in coordination with mflores from IT multiple ideas for improvement were discussed.

Ideas¶

• IT shared responsibility: DNS should not a problem as DNS was already moved out of specific VMs to centrally maintained instances. With that we do not have ssh access to any DNS server. If fozziebear still runs DNS according to mflores then IT can move DNS to a central instance as well. We do not have ssh access to fozziebear
• We all want to have a consistent setup. Right now walter+walter+fozziebear are inconsistent. Regarding DHCP according to mflores other teams also use DHCP relay. QE wants to have at least log access to DHCP servers but we do not necessarily need root write access over ssh . mflores explained that there is already log aggregation and forwarding done for other purposes. that could be extended for DHCP logs . Related issue https://jira.suse.com/browse/ENGINFRA-3732
• IT does not want to maintain PXE. We would be fine to maintain PXE servers if VMs are provided by EngInfra. It should be possible to manage VMs in openplatform including a PXE server. We can request a specific network for openplatform and then create VMs in those networks