action #33466
closed
coordination #34858: [functional][sle][opensuse][y][epic] Ensure HDD sizes are consistent and have a reasonable relation to product specifications
[functional][osd][y][yast] The size of disk used for 64bit and aarch64 machine test are not the same, causing different behavior to handle per architecture
Added by SLindoMansilla about 6 years ago.
Updated almost 5 years ago.
Category:
Bugs in existing tests
Target version:
SUSE QA - Milestone 25
Description
Observation¶
The size of disk used for 64bit and aarch64 machine test are not the same, causing different behavior to handle per architecture.
Here an example:
At least in this case, the size of disk comes from the machine:
My suggestion is to also use 20GB on aarch64 to make test behaves the same as 64bit.
Further details¶
Always latest result in this scenario: latest
[19 Mar 2018 15:04:03] SergioAtSUSE: yes, we should use the same default unless explicitly stated in the test suite. This is also why I was against just bumping the HDD size to 40GB in all cases where we failed to find btrfs subvolumes
[19 Mar 2018 15:05:12] SergioAtSUSE: what I see looking at https://openqa.suse.de/admin/machines is a wild mix with no proper comments or anything there. Better create a ticket for that and try to find some more references in archives first before changing anything. This looks like a bigger issue
[19 Mar 2018 15:05:55] for example why the heck should svirt-kvm be 24GB?
[19 Mar 2018 15:06:41] and s390x-kvm-sle12 and s390x-kvm-sle15 has 10GB
[19 Mar 2018 15:07:09] rbrown: do you remember anything why we have 20GB for some openQA machine definitions and 10GB for others like aarch64?
[19 Mar 2018 15:07:44] 10GB is the default in os-autoinst IIRC so the question is mainly "should we explicitly set at least 20GB in all cases?" first
[19 Mar 2018 15:08:08] okurz, the aarch64 workers all had insufficient diskspace to run 20GB workers
[19 Mar 2018 15:08:40] hm, ok. the same might apply to s390x in some cases but that does not make our testing results easier to understand :(
- Subject changed from [functional][osd] The size of disk used for 64bit and aarch64 machine test are not the same, causing different behavior to handle per architecture to [functional][osd][y][yast] The size of disk used for 64bit and aarch64 machine test are not the same, causing different behavior to handle per architecture
- Target version set to Milestone 17
- Target version changed from Milestone 17 to Milestone 21+
- Target version changed from Milestone 21+ to Milestone 21+
- Parent task set to #34858
- Target version changed from Milestone 21+ to Milestone 23
- Due date set to 2019-03-26
- Due date changed from 2019-03-26 to 2019-06-18
- Target version changed from Milestone 23 to Milestone 25
Let's keep at is until SLE15SP1 is released.
- Status changed from New to Resolved
- Assignee set to riafarov
Fixed in the test suite settings. I've checked couple of other scenarios, e.g. RAID0 and problem is not there.
Also available in: Atom
PDF