QA

https://jira.suse.com/browse/ENGINFRA-3744 was set to "Done" but I have no full confirmation that the machine is usable as in before.

@dawei_pang can you confirm that haldir.qe.prg2.suse.org is fully usable? The machine is within the HMC https://powerhmc1.oqa.prg2.suse.org/ and you have access there so you could use that access.

Let's use the machine as openQA worker like redcurrant

Will need to clarify with acarvajal who mentioned haldir on https://confluence.suse.com/display/qasle/QE-SAP+Power9+Infrastructure

https://suse.slack.com/archives/C02CANHLANP/p1712665391961199

@Alvaro Carvajal in https://confluence.suse.com/display/qasle/QE-SAP+Power9+Infrastructure you mentioned haldir but previously when I asked I got the answer that QE-SAP doesn't use haldir. So should we prepare haldir as a generic openQA PowerVM host or do you use haldir within QE-SAP? Context: https://progress.opensuse.org/issues/153718

Got confirmation from acarvajal that haldir is free

yes. that confluence page is WIP and the haldir part was taken (copy & paste) from the old document at https://gitlab.suse.de/hsehic/qa-css-docs/-/blob/master/infrastructure/power9-configuration.md. I asked internally and haldir should be free

Recovered the password for the VIOS padmin-user and reset it to the default. https://gitlab.suse.de/OPS-Service/salt/-/merge_requests/4976 created to give the VIOS a basic working network config. Alvaro made me aware of https://sd.suse.com/servicedesk/customer/portal/1/SD-153996. Next I try to setup a working RMC connection to the HMC to be able to configure a virtual network+disk for the LPARs and connect them with each other.

managed to figure out which interface is connected and already statically configured haldir-vios4 which is able to be reached:

workstation pillar/domain ‹add_haldir› » ping 10.145.0.103
PING 10.145.0.103 (10.145.0.103) 56(84) bytes of data.
64 bytes from 10.145.0.103: icmp_seq=1 ttl=253 time=218 ms
^C
--- 10.145.0.103 ping statistics ---
2 packets transmitted, 1 received, 50% packet loss, time 1001ms
rtt min/avg/max/mdev = 217.517/217.517/217.517/0.000 ms

I now hit the same issue as described in https://sd.suse.com/servicedesk/customer/portal/1/SD-153996 and will follow that one. But currently we can't do anything more from our side.

SD ticket resolved and the RMC connection works now. I will continue with a network and disk configuration before defining the SUT LPARs.

Setting due date based on mean cycle time of SUSE QE Tools

https://gitlab.suse.de/openqa/salt-pillars-openqa/-/merge_requests/782 which can already be merged but only tested after the OPS-request is merged as well.

started to validate the instances but realized that the OPS request is still not merged. Asking in #dct-migration
Also for later my command to create validation jobs:

for i in {1..10}; do echo openqa-clone-job --within-instance https://openqa.suse.de/tests/14092616 --skip-chained-deps --skip-download TEST+=-poo153718#${i} BUILD=nsinger_validate_poo153718 _GROUP=0 WORKER_CLASS=hmc_ppc64le_poo153718; done

All jobs reached PXE and even loaded kernel+initrd silently fail and fall back into grub. Not sure yet what causes this, have to investigate.

Machine fails to load kernel+initrd over PXE: https://openqa.suse.de/tests/14115017#step/bootloader/28
We agreed to use cloudberry (https://progress.opensuse.org/issues/153727) as reproducer to see if this is a common problem between these machines or if it is specific to haldir.