dockerdata-nfs data size causes ONAP crash #so #dublin #mariadb #nfs

Ferrero Marco Angelo

Hi everybody,

In the last two months, we have being used ONAP executing the vDNS-vLB use case many times.
Yesterday  ONAP suddenly stopped working.
According to the result of our analysis it happened because there was no free disk space on the VM hosting the shared directory /dockerdata-nfs .
More specifically such directory is hosted on the VM nfs-server which was deployed with 200GB of disk space used in this way: OS and other stuff occupies 6GB while the remaining 194 GB are occupied by the dockerdata-nfs directory.
Today we have re-installed Dublin from scratch and we notice that /dockerdata-nfs directory occupies 15GB

We have the following concern:
Is this an ONAP design issue or did we configure it in a wrong way?
….and the following questions:

Has anyone else experienced the same issue?

What are your/dockerdata-nfs directory occupation statistics ?

Are there some scripts or whatever that could be used in order to reduce the amount of data in /dockerdata-nfs directory (for example by deleting some older logs ...) to avoid the system to crash in the future?

Thank you
Best regards,

Marco Ferrero