Created attachment 1013179 [details] vdsm log Description of problem: After trying to import an NFS-based export domain from an empty path, a stale connection to that path prevents creating a new domain there Version-Release number of selected component (if applicable): vdsm-4.17.0.3-176.git8e5a63e.fc20.x86_64 ovirt-engine-3.5.1.1 How reproducible: Always Steps to Reproduce: 1. In the UI, attempt to import an export domain from an empty NFS remote dir 2. After engine reports an error, use the UI to attempt to create a new export domain at that same path Actual results: Engine reports the error: Error while executing action: Cannot add Storage Connection. Storage connection already exists. Expected results: The export domain should be created successfully. Additional info:
Created attachment 1013180 [details] engine log
Daniel/Maor, iirc, this issue is already tracked in a different BZ?
(In reply to Allon Mureinik from comment #2) > Daniel/Maor, iirc, this issue is already tracked in a different BZ? yes, it is a duplicate candidate for https://bugzilla.redhat.com/1136930. I've tried to reproduce it as Adam described, but it looks that it works on my env. 1. Tried to import an empty export Storage Domain 2. Got the following error: "There is no storage domain under the specified path. Check event pane for more details" 3. Validate - Checked the DB table of storage_server_connection - It did not contained the relevant path 4. Try to add a new NFS Storage Domain of the same path and succeeded Adam, have you used oVirt 3.5.1.1? I see that the fix was already merged for oVirt 3.5.1 so it should be there, unless my reproduce steps are different then yours... was the error you got when importing the export domain was the same as mine, "There is no storage domain under the specified path. Check event pane for more details"?
I am still using an old oVirt-3.5.0 build. If you guys already addressed this issue in a different bug I am comfortable marking this one as a dup. If the issue persists in 3.5.1.1+ than that first bug can be reopened.
Closing based on that comment. If we see this problem return in a newer build, we can always reopen the original bz. *** This bug has been marked as a duplicate of bug 1136930 ***
(In reply to Allon Mureinik from comment #5) > Closing based on that comment. > If we see this problem return in a newer build, we can always reopen the > original bz. > > *** This bug has been marked as a duplicate of bug 1136930 *** I am still affected by this bug in 3.6.1