Just in case if it helps others, Answer from salesforce support
R&D and Tier 3 have provided a root cause analysis:
A single organization's storage might be spread across multiple tables for load balancing. Because we're a shared hosting service, those individual tables will also contain data from other organizations. This means that each of your individual objects might be saved to different tables, and those tables might contain data from multiple organizations.
The storage for the table that Generic_Log__c was reaching had exceeded its storage, resulting in the error you encountered.
There is an automated job in place that monitors for space usage and adds storage once it has reached a certain threshold. In this case, a bug was encountered that prevented this automated increase from happening.
Since the storage exists and the job processes at a table level on the instance and not an individual organization level:
- Multiple orgs might have been impacted if they attempted to save data to the impacted table.
- A single org might be impacted on one object that saved data to the impacted table, but not on another object if it saved data to a different table.
- A single org might not have been impacted at all if that org did not save data to the impacted table.
A bug was created to address the issue with the automated job. R&D has identified a fix for the bug, and we're waiting for confirmation on whether it is still pending deployment or has been deployed already.
There was an additional question provided about why did the record save process work for a time. It's hard to say at this time, but it's likely that the storage was teetering on the "exceeded" limit at that time, and data was deleted in some way (example, sandbox refresh) that temporarily brought the storage below the limit.