Hi Matej,
{slice=0.0, objectID=5714869, requestID=5001010, sliceDate=2017-05-06T00:00:00}
The items marked in bold are causing the conflict (on the basis a record in prj_blb_slices already exists with the same values).
Possible causes I suspect may include:
You somehow have a duplicate request, make sure this returns only 1 record: select * from prj_blb_slicerequests where id = 5001010
You have a duplicate source record. Since Slice Request ID 5001010 is custom, I don't know what it points to. Can you check? Time entries, assignments, actuals, etc. Once we know what that is, we can then: select * from <src table> where prid = 5714869
We want to make sure that too only returns 1 record. If the slice request is for time entries, then make <src table> prtimeentry and if it's estimates/actuals make it prassignment. If it's allocation, make it prteam, and finally if it's availability, then make it prj_resources. If your slice request is on anything else, we'll need to check further.
If both of those only contain 1 record then possibly an issue with your prj_blb_slices table (or its indexes) are having a bad reference in them, as the records should be getting deleted from there before the slice job inserts replacements.
You could try deleting this range of records as per the index definition explicitly (actually I'd probably do all records for the slice request id and object id, regardless of date), like this: delete from prj_blb_slices where slice_request_id = 5001010 and prj_object_id = 5714869
Make sure to COMMIT afterward.
And if that doesn't work, then I'd suggest a full re-slice of that request, which you can effect by going to the time slice definition in the UI and making a change (e.g. change the number of periods), save it, then make another change to put it back again. The Last Run Date for the slice request should then clear and a full reslice will take place next time.
If after this, there are still errors, it may be required to do something about getting a new backup exported/imported, but check via a support ticket first if possible and reference these steps/actions that you already undertook.
Hope that helps you out.