We've just changed the ZAP core to handle database / diskspace limits a little bit more cleanly.
Previously if ZAP hit either of these problems then it would just keep logging errors.
Now if you are using the UI it will popup a warning dialog.
If you are using ZAP in automation then ZAP will exit, with a suitable error message.
Based on our testing it does look like ZAP can usually write to the zap.log file even if we fail to get more diskspace allocated to the database, but we also now write the warnings to stderr.
We are also looking at handling memory issues as well, but those are being implemented in an add-on which we expect to be included in 2.17.
The core space checking changes have been merged and so will be in the next nightly and weekly releases.
We do not anticipate that they will cause an problems, as if you run out of space then ZAP would have failed before anyway, but if you do see a change in behaviour in ZAP then please let us know ASAP.
Many thanks,
Simon