![]() This could be solved by not using a database, or at least using a much simpler database structure. This is primarily a speed issue (and there are some recreate errors as well). Then there is the problem with recreating the local database. The errors that produce an inconsistent database should really be found and fixed asap. Again, removing the database will not fix these, just make them show up later on where they are more likely to prevent correct restores. Removing the database here will just make the problem invisible until you really need the files.Īnother problem with the database is consistency issues. In some cases there are failures because the remote store reports different contents (for whatever reason). However, many of the issues will not be fixed by simply removing the database. In Duplicati, there are quite a few problems related to the database. Theoretically, it should be much faster to run queries agains a database, and it is also more crash resistant. It’s a very “sophisticated” approach, and though elegant, it creates several points of failure. I think databases use is the weak point of Duplicati. ![]() ![]() With only 1 thread running, it is still writing at over 130MB/s for separate processes.īased upon this, I am guessing that an SSD would actually benefit far more from multiple threads, because there is no seek involved.įrom: Duplicacy Issue: optimization of speed (was: one thread for local HD backup) The array normally writes at 200MB/s, but with more than 4 backup threads running, it slowed down to less than 9MB/s. In fact, the way I pinpointed what was going on is running a separate speed test on the drive while the multi-threaded processes were running. That’s because writing each chunk requires a head seek, and if there are multiple threads, that’s a LOT more seeks going on. In thinking about it, this may make more sense than it seems for a spinning hard drive. The bottom line is that only one clear winner emerged: 1 single backup thread. threads 32 ~2.0MB/s - This was a longer test than for 16 threads, and converged to a somewhat higher rate, though not significantly different than 2 threads threads 16 ~619KB/s - seems even slower, but I did not let this one run very long, so it may not have stabilized at full speed threads 64 ~1.1MB/s - extremely slow, earlier on before it stabilized, it was running at only 100KB/s threads 2 ~1.93MB/s - very slow, it was slated to take well over a week threads 1 ~12MB/s - Not great, but at least it will finish the initial backup in about a day or two. It is not rigorous, but at least an indicator. By informal I mean that I would kill the ongoing backup, then resume at the same spot with a different thread count, wait until it stabilized on a long stretch where there were minimal duplicates, and then recorded the MB/s. Looks like this is not necessarily the case: I will also have to note, that CY can perform even better with multithreaded upload. So the difference to duplicati is not as big, but duplicati still seems better here. Search also works in the GUI version, but you can’t search across revisions. You can only do it via CLI ( as described here) and even then, you have to run a separate command to then restore the identified file. I’m inclined to add this as another plus for duplicacy.Įdit2: I just discovered the very (!) basic restore functionality of duplicacy: it is very cumbersome to search for and restore a file whose exact location you don’t know. unable to do cross-source deduplication (which, in my use case, is pretty nifty to have as it allows you to move large files or folders all over the place locally, without having to worry about that making your backend storage explode).Įdit: I just discovered that duplicacy allows you to change the password for your archive.On the minus side, duplicati is (currently) Personally, I really like the Web UI and the way it makes command line options available without forcing you to use the CLI. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |