Problem upgrading from 14.0.3 to 14.0.4


Just tried updating from 14.0.3 to 14.0.4 and getting an exception error, tried both from web and occ command:

2018-12-13T16:48:11+00:00 Repair step: Repair pending cron jobs
2018-12-13T16:48:11+00:00 Repair info: No need to repair pending cron jobs.
2018-12-13T16:48:11+00:00 Starting code integrity check…
An unhandled exception has been thrown:
TypeError: hash() expects parameter 2 to be string, boolean given in /home/ncsave/public_html/lib/private/IntegrityCheck/Checker.php:209

Nextcloud version (eg, 14.0.3):
Operating system and version (Debian 9):
Apache or nginx version (eg, Apache 2.4.25):
PHP version (eg, 7.0.33):

The issue you are facing:

Cannot complete update, can go out of maintenance but asked again to update …

Is this the first time you’ve seen this error? (Y/N):Y


Try this.

sudo -u www-data php occ files:cleanup
sudo -u www-data php occ app:update
sudo -u www-data php occ check
sudo -u www-data php occ upgrade

Thks for your answer. As I said, I already tried the occ method to no avail.
After much looking around and testing I found what the problem was. My instance is on a virtualhost and data on an external attached drive (OpenStack Public Cloud). At one point I saw that much higher in the logs I had a write error due to disk quota and stupidly I spent quite a long time trying to understand why my data disk was full: from Debian CLI, I was seeing a 45% occupation and a very reasonable inodes occupancy as well … I say “stupidly” because in fact I had reached a quota on the virtualhost itself, and upgrading was in fact pushing nextcloud just over the limit which was quite low, 1gb. Of course I had no memory setting up that limit, I guess that was a default somewhere that I kept without noticing … That installation has been up and running for almost a year, has been updated from 12 to 13 to 14 and is handling close to 500gb of data with no problem so far, just that last minor upgrade did push it over that limit :frowning:
Well anyway, thanks for your kind answer and sorry for the noise, it was all my fault !


If I may suggest, this is something that I am doing on my NC servers, just to keep an eye on things.
Install Glances, InfluxDB and Grafana to monitor your activities, keep the monitor open at all time and if anything it will show.
This has saved me on many occasions, especially with MEM and the CPU usage.

Here is an example, this is monitoring activities for ~15 current active users at the time of capture