The integration testing definitely isn't just your environment and is
certainly from the change to alpine as a base image. It looks like Koji
reported some issues around the Windows service installation/support which
I think is important to make sure we get right, so I will be cancelling
this
Jim,
To be honest, I would recommend making those changes directly to the existing
PublishAMQP processor, as unless I am missing something, that does not seem
like a breaking change that would affect other users negatively. I’d suggest
opening a Jira for the feature, cloning the GitHub
One Strange thing one of our testers found was that we're using the default
12hours archive and 50% disk space configuration, he noticed that when nifi
removed the archived files inside the content_repository, he checked the
inodes ' count didn't go down, then he tried to remove some of the
Virgil,
This was intentionally introduced via NIFI-3907 [1] in Apache NiFi 1.3.0 as a
mitigation for CVE-2017-7667 [2]. Prior to this change, a malicious site could
have displayed the NiFi UI and introduced invisible overlays such that an
unsuspecting user would perform actions like entering
Hi Michael, the no space left on device occurred again and I checked the
inodes at the time and found it was indeed full, why would the inodes
become full and are there any solutions to get around this problem? Thanks.
/Ben
2017-12-13 13:36 GMT+08:00 尹文才 :
> Hi Michael, I