Hi,
On 6/16/21 7:28 AM, Yedidyah Bar David wrote:
On Sun, Jun 13, 2021 at 10:30 AM Yedidyah Bar David
<didi(a)redhat.com> wrote:
> Hi all,
>
> Last two nightly runs of [1], 2052 and 2053, failed both after
> finishing all tests, during collecting artifacts, with:
>
> 06:31:39 @ Collect artifacts:
> 06:31:40 # [Thread-1] lago-he-basic-suite-master-engine:
> 06:31:40 # [Thread-2] lago-he-basic-suite-master-storage:
> 06:31:40 # [Thread-3] lago-he-basic-suite-master-host-0:
> 06:31:40 # [Thread-4] lago-he-basic-suite-master-host-1:
> 06:31:40 Unknown exception: q must be exactly 160, 224, or 256 bits long
The nightly job keeps failing like this, and also the node-upgrade suite
fails the same way.
Any idea?
Might be related to changes around artifacts collection?
It's caused by the
fact that we've switched to using FIPS in OST.
We already had a discussion about it - it was about paramiko
not properly supporting the RSA keys with SHA256/SHA512 checksums.
There's no point on working on this - we should focus on switching
to ansible-based artifact collection in these suites.
If that causes CI-1 votes, then let's disable the old artifact collection
in respective 'control.sh' files.
Regards, Marcin