I can not even push to my own fork:

$ git push praiskup chroot-blacklist 
git@pagure.io: Permission denied (publickey).
fatal: Could not read from remote repository.
Please make sure you have the correct access rights
and the repository exists.

I can not even push to my own fork:

Does this still happens? Looking at the gitolite config, this should work fine.

I can not merge this:
https://pagure.io/copr/copr/pull-request/432
After click on "merge" button I am only informed about the fact that "The pull-request can be merged with a merge commit".

Could you share a screenshot of this?

Does this still happens? Looking at the gitolite config, this should work fine.

Yes.

Could you share a screenshot of this?

Interesting, I reviewed my list of keys uploaded to pagure - and there seems to be only my pretty old ssh key (so I started using that again and it's fine now). Is it possible that pagure cleaned my uploaded keys itself, somehow?

Btw, I was able to push one commit yesterday, and in about ~2 minutes I tried to push another one and I was suddenly blocked. So something really happened on background.

But... I resolved the "pushing to my fork" issue, and pushing to project I administrate. The merge button still doesn't work.

Also note this:

remote: Running hooks for copr/copr
remote: No handlers could be found for logger "pagure.hooks.default"
remote: Traceback (most recent call last):
remote:   File "/usr/lib/python2.7/site-packages/pagure/hooks/files/git_multimail_upstream.py", line 2650, in <module>
remote:     main(sys.argv[1:])
remote:   File "/usr/lib/python2.7/site-packages/pagure/hooks/files/git_multimail_upstream.py", line 2621, in main
remote:     recipients=options.recipients,
remote:   File "/usr/lib/python2.7/site-packages/pagure/hooks/files/git_multimail_upstream.py", line 2567, in choose_environment
remote:     "EffectiveEnvironment", tuple(environment_mixins) + (Environment,), {}
remote: TypeError: type() argument 1 must be string, not unicode
remote: Traceback (most recent call last):
remote:   File "/usr/lib/python2.7/site-packages/pagure/hooks/__init__.py", line 373, in run_project_hooks
remote:     changes=changes,
remote:   File "/usr/lib/python2.7/site-packages/pagure/hooks/__init__.py", line 111, in runhook
remote:     changes=changes,
remote:   File "/usr/lib/python2.7/site-packages/pagure/hooks/mail.py", line 113, in post_receive
remote:     raise Exception("git_multimail failed")
remote: Exception: git_multimail failed
remote: Running hooks for copr/copr
remote: Sending to redis to log activity and send commit notification emails
remote: * Publishing information for 1 commits
remote:   - to fedmsg
remote: git_multimail failed

Interesting, I reviewed my list of keys uploaded to pagure - and there seems to be only my pretty old ssh key (so I started using that again and it's fine now). Is it possible that pagure cleaned my uploaded keys itself, somehow?
Btw, I was able to push one commit yesterday, and in about ~2 minutes I tried to push another one and I was suddenly blocked. So something really happened on background.

Ah, I have an idea, did you by any chance had your ssh key in two different accounts?
Yesterday we upgraded from 5.0 to 5.1 which brought some change in how ssh keys are stored in the database. Most of the migration went fine but since we now enforce uniqueness on the SSH key, if there was a key present multiple times it may have lead to some odd situation like this one.

But... I resolved the "pushing to my fork" issue, and pushing to project I administrate.

Cool

The merge button still doesn't work.

This worries me a little, going to try to look into it more

Also note this:
remote: raise Exception("git_multimail failed")
remote: Exception: git_multimail failed

Could you open another ticket for this one please?

Ah, I have an idea, did you by any chance had your ssh key in two different
accounts?

That is possible. I'll review later.

Also note this:
remote: raise Exception("git_multimail failed")
remote: Exception: git_multimail failed

Could you open another ticket for this one please?

3887

This ticket and #3886 have the same root cause, which I just found and fixed. I need to add some tests for my fix and I'll send a PR

Commit 2a1d4db8 fixes this issue