QA (public) » openQA Project (public) » openQA Infrastructure (public)

I thought the machine would not live in a server room but apparently I was wrong.https://racktables.nue.suse.com/index.php?page=object&tab=default&object_id=742 is the racktables entry. I created https://infra.nue.suse.com/SelfService/Display.html?id=162247

Thanks to gschlotter the machine is back. He reinstalled it – multiple times as we misunderstood regarding requirements – but we are back with a transactional server. Current setup:

transactional-update pkg in openQA-worker htop gdb os-autoinst-distri-opensuse-deps
chown -R _openqa-worker /var/lib/openqa/{cache,pool}
reboot
systemctl enable --now openqa-worker-cacheservice openqa-worker-cacheservice-minion openqa-worker@1

cat /proc/mdstat

md0 : active raid0 sdc2[1] sdb2[0]

lsblk

NAME        MAJ:MIN RM  SIZE RO TYPE  MOUNTPOINT
sda           8:0    0  3.7T  0 disk  
sdb           8:16   0 29.8G  0 disk  
├─sdb1        8:17   0    1G  0 part  
└─sdb2        8:18   0 28.8G  0 part  
  └─md0       9:0    0 57.6G  0 raid0 
    └─md0p1 259:0    0 57.4G  0 part  /
sdc           8:32   0 29.8G  0 disk  
├─sdc1        8:33   0    1G  0 part  [SWAP]
└─sdc2        8:34   0 28.8G  0 part  
  └─md0       9:0    0 57.6G  0 raid0 
    └─md0p1 259:0    0 57.4G  0 part  /
sr0          11:0    1 1024M  0 rom   

As the machine has a big but slow HDD but pretty much RAM I think previously we used a ramdisk for the openQA worker pool, e.g. we could do echo 'tmpfs /var/lib/openqa/pool tmpfs nodev,nosuid,noexec,nodiratime,size=64G 0 0' >> /etc/fstab but since we want to use the same filesystem for cache+pool I don't see this working out. For now I will configure the machine without ramdisk and without use of the HDD with a single worker instance as there is currently 40GB free.

EDIT: 2020-02-11: … which was obviously a bad idea because we need another 50GB by default for caching. I now configured /dev/sda1 with an ext4 that I mounted on /var/lib/openqa.

The worker instance imagetester:1 executed jobs just fine over the past day. Enabled a second worker instance.

looks good now