Very large dataset, running out of memory

[mohuddle]

I'm running licensed BC4 comparing a local partition and a networked partition. Scan seems to run fine and has made it through quite a few directories. However, when I leave it and check back after a day or so, the app has either died or it will throw an 'out of memory' error. I've noticed the app itself, on this particular scan, will eventually consume all 16G of memory available.

Are there tweaks to help it not consume all the memory? I've tried to set the compare using directory name wildcards, but it doesn't seem to support this.

Any suggestions?

[Zoë]

In the Session Settings/Rules dialog you can use filters to limit the scan to a subset of directories.

Select the Session->Session Settings... menu item (or the Rules toolbar button), and on the Name Filters tab, change the '*' in "Include folders". The documentation for the wildcards you can use is here. If you use something like .\[a-m] it will only include the top level folders that begin with 'a' through 'm'. You could then do a second compare with .\[^a-m] or .\[n-z] to catch all of the rest of the directories.

Normally BC has a single copy of itself that runs for all of the comparisons. If you run it with the /solo command line switch, it will stay separate, which would allow you to do multiple comparisons without them all running out of room.

Unfortunately we don't have any other settings that would adjust how much memory BC is using. Since the primary user interface is interactive comparisons and syncs, rather than set-it-and-forget-it, we have to keep a lot more things in memory.

[mohuddle]

Thanks Zoë, that will help. I can limit my folders to A*\, let that finish, B*\, and so on.