How to recover a backup set after hw failure?

Depending on the message type it should be available.

Messages that just say “Result” are supposed to be summaries (with the three dots) of OTHER messages. Those other messages should be fully viewable in one of the 4 different log areas (yes, 4).

I just checked on 2.0.3.6 and I thought every log had a “Load older data” button at the bottom but right now I’m only seeing it on the “Remote” log page for specific jobs.

@kenkendk, am I mis-remembering or did all logs used to have a “Load older data” button?

I already commented on this. On windows, I only see Results in the log listing, but Ubuntu, I always see Message lines between Result lines, regardless of the success or failure of the run. Moreover, all look alike:

Kuvakaappaus%202018-06-10%2008-23-30

From that kind of listing, I cannot see if there are errors or warnings included, they all look the same. This is a big part of the problem. I had been looking at these long lists of messages for so long that when warnings and errors did begin to appear, I had already trained my eyes to skip all the message lines. If, however, the warnings were marked as warnings, and errors as errors, in this listing, I would have noticed the difference. Since they are all marked messages, I have to open the message like this:

then open up the Result line, and look inside, to find out if this is an error or warning or an ordinary message. If it happens to be one of the messages behind the three dots, I cannot be sure of the status.

This is not helpful. If the behaviour was what you say, if the messages only appeared only when they actually are warning or errors, it would indeed be easier to notice that something is wrong, but when every successful run produces loads the these messages, and warnings are errors are all labelled messages, this is more of a nuisance than helpful.

Also, asking me to “please verify the sha256 hash” is not helpful. When the files are on my external drive, and I can see they are zero size, there is no need to verify the signature. Instead I need advice on how to proceed to fix the problem.

I am still waiting for advice on what to do with the failed purge-broken-files operation.

You might want to start a new topic in #features to discuss your messages / results / warnings / errors suggestions.

As for purge-broken-files issue, your post here seems to indicate that the purge step was barely even started before it got interrupted by the “Unexpected number of remote volumes marked as deleted” error.

Have you tried a database repair (no recreate) since that “remote volumes” message appeared? I suspect there might be an issue where a repair fixes the “first level” of issues it finds but if there are secondary (tertiary, quaternary, etc.) issues they don’t get resolved until the repair is re-run.

Once those issues have been addressed, the purge-broken-files should be able to proceed.

Unfortunately, I got distracted and did not follow up. I ended up scrapping my failing backup job, deleting all its files, and re-creating the job. After all, I was running Duplicati as a secondary backup solution, not the primary.

Since then, however, I have had bad luck with the USB connector, and even forgotten that my backup was running and pulling it out prematurely myself, and I can tell you that it has never again caused me any problem; Duplicati has always been able to recover from it without a hitch. Most of this time, if not all the time, I have had Duplicati 2.0.4.5 beta. So I am happy that the problem (hopefully) got fixed way back, even though we never got to the bottom of it.

Glad to hear things are working more smoothly for you now!

Wong thread. It doesn’t help you spreading the same message all over without thinking.