I ran into this same issue on a CUCM 11.5 system on May 14, 2019. The conditions and error message of "/usr/local/platform/bin/startcliscript.sh: fork: retry: Resource temporarily unavailable" were the same. I ended up having to reboot the CUCM Publisher I was receiving this message on. After the reboot I was able to log into the CLI successfully. Several days ago, I had added the credentials for the CLI access account to the IR Prognosis tool we recently acquired for system monitoring. I'm wondering if Prognosis had established an SSH session and was causing this issue.