[ZODB-Dev] Checking FileStorage integrity

Chris Withers chrisw@nipltd.com
Wed, 08 Aug 2001 17:24:40 +0100


Michel Pelletier wrote:
> 
> Just as an experiment, you could try iterating over all objects and
> poking them into a fresh storage.  Log raised exceptions, etc.

I'm pretty sure that's what copyTransactionsFrom does.
To be on the safeside, I'm doign a .zexp of the whole lot and I'll re-import it
when it's done. It's been all day and it's done 1.4G out of 1.8G. Maybe it'll be
done by the tiem I get in tomorrow ;-)

I'm also less sure that there's a problem with the ZODB here, I removed the
LeakFinder product and a couple of other things which cleaned up some problems
on the development machine.

Mindyou, I absolutely cannot seem to get LeakFinder to die :-( I've deleted it
from the Product tab, I've used the following external method to get rid of it
from the Control Panel:

def fix_cp(self):
    cp = self.Control_Panel
    try:
        del cp._objects
        return "done"
    except: pass
    return "arsed"

...and obviously I'v deleted the product's folder on the file system.

Nevertheless, I still get the following whenever I start up Zope:

------
2001-08-08T16:16:37 PROBLEM(100) ZODB Could not import class 'LeakFinder' from
module 'Products.LeakFinder.LeakFinder'
Traceback (innermost last):
  File /usr/local/zope/2.4.0_base/lib/python/Zope/ClassFactory.py, line 98, in
ClassFactory
ImportError: No module named LeakFinder.LeakFinder

If anyone knows where Shane is, I could really do with a hand on this one ;-)

cheers and thanks for the suggestion,

Chris