Auto decrypt not working

darlo

New Member
Hello All,
I have notice that since the 13th July none of my recordings are being decrypted. I have done a few things like confirm content share is still on, disabled and re-enabled the recursive decrypt flag with no change. Files are being added to the queue for decrypting but not being decrypted.
Looking at the auto log I can see the following error "decrypt::run: database disk image is malformed" and I am now at a loss for what to do next.

Can anyone advise. Please keep it simple I am very much a novice here

Many Thanks

Ian
 
... none of my recordings are being decrypted. I have done a few things like confirm content share is still on, disabled and re-enabled the recursive decrypt flag with no change. Files are being added to the queue for decrypting but not being decrypted.
Looking at the auto log I can see the following error "decrypt::run: database disk image is malformed" and I am now at a loss for what to do next.
The error means that one of the SQLite databases used in the system is corrupt, perhaps the queue database /mod/etc/queue.db. An auto-processing guru should be able to help.

Meanwhile, can you decrypt "manually", ie select the recording in Webif>Browse Files and then Opt+>Decrypt>Perform decryption ?
 
I have just tried to manually decrypt and it sort of works. Once the process has finished I get a copy of the original file with a .decrypting extension. Which has been decrypted but cannot be seen by the humax until I rename it to .ts
 
A bit more info the line in the auto log immediately prior to the disk image msg is always the same (and the same time) so taking the most recent I have

4832 16/07/2019 20:48:52 - /mod/webif/lib/queue.class:144 @
4831 16/07/2019 20:48:52 - ::decrypt::run: database disk image is malformed

after this it stopped trying to decrypt as it was recording at 9.
 
Files are being added to the queue for decrypting but not being decrypted.
Does viewing the queue still work - usually when the queue database is corrupt you cant view the queue and nothing is added to it.
It wouldn't hurt to rename or delete /mod/etc/queue.db to see if it resolves the problem since it will be recreated on next use
 
I have renamed the queue and it was pretty quickly updated with a list of all the items needing to be decrypted. Had to wait until the box had completed recording but it is currently going through the items and decrypting them

Many thanks for the help
 
Just to wrap up, since @MymsMan provided the solution.
A bit more info the line in the auto log immediately prior to the disk image msg is always the same (and the same time) so taking the most recent I have

4832 16/07/2019 20:48:52 - /mod/webif/lib/queue.class:144 @
4831 16/07/2019 20:48:52 - ::decrypt::run: database disk image is malformed

after this it stopped trying to decrypt as it was recording at 9.
And that confirmed that the queue DB was the problem (the 2 lines are part of one message). But the system appears to have failed while trying to add a recording to the queue for decryption rather than when dequeuing for decryption, through this line (30) at /mod/webif/lib/auto/plugin/decrypt/auto.hook:29-31
Code:
        # Enqueue file                                      
        queue insert $ts decrypt
        ::auto::log "  $file - Queued for decryption."
which invokes the SQL operation logged above at /mod/webif/lib/queue.class:144
Code:
        $db query {      
                insert or ignore into queue(submitted, file, action, status)
                values(%s, '%s', '%s', '%s')
        } [clock seconds] $file $action $status
So does the auto processing stop completely after this error, resulting in no further queue operations at all, including decryption?
 
Last edited:
The only items I ever see in the queue are decryption (I think). What else is normally in there?
 
I don’t think I would see anything else in the queue then. The only other thing I use is sweeper
 
I've just noticed that none my recordings have worked since July 19th. Looking in Queued Tasks I see that the box is permanently trying to decrypt one recording. I cannot delete this task through the Queued Tasks interface. So I've Telnetted in to try to delete the queue db as above but the file system is read only. This must be such a newbie question, for which I apologise, but how can I rename/delete this file? Thanks.
 
I am not sure on using telnet. I logged into the root directory using filezilla (user "root" and normal password) and was able to rename the file from there.
 
Ah well, turning it off and back on again has made the file system read-write so I can now delete the queue.db file. (Hoping that's fixed the problem).
 
Had your reboot not sorted the read/write, you would have needed to run fixdisk from the Telnet maintenance mode menu.
 
Ah well, turning it off and back on again has made the file system read-write so I can now delete the queue.db file. (Hoping that's fixed the problem).
It is possible/probable that it was the disk being read-only that was causing the decrypt to fail and that the queue.db was an innocent victim.
Fortunately deleting it is normally harmless with most files being requeued automatically.
 
Thanks for your replies, everyone. Turning it on and off didn't fix the problem - it did allow me to delete the queue.db file but that didn't allow recordings to be made. After further tinkering I discovered that existing recordings were still read-only and that led me to find a post here somewhere which pointed me at FixDisk (cudos, Black Hole!!)

Am now running FixDisk - sda1 and sda3 were fine (and fast) but it's found and fixed a couple of things on sda2 (slow) so far. I imagine sda2 is where recordings are stored.

Hopefully it'll sort itself out.
 
Back
Top