Monday, September 19, 2011

Cache problem on Bridge CS4

As digital photo librarian one of my tasks is to keep Bridge on my four machines constantly indexed and updated: three have admin level and one a simple user level. I call these my benchmark machines since they are the ones by which I measure all the other users' machines in the department. Typically, they the most accurate and up-to-date for searching; they include two PCs and two Macs.

Back in July of this year, Bridge on one of my two benchmark Macs displayed the feared "purge the central cache" message:


There is no workaround and no alternative but to purge the central cache and completely reindex the machine from scratch.

Last Friday the same thing happened to my other benchmark Mac, a laptop.

And no, we don't know why this happens: the cache size remains at normal levels:44 gbs with 380k+ items (July) and 40 gbs with 306k+ items (September). No unusual activity at the server end and so far no one else in the department has experienced this problem. Of course, that last observation leads me to suspect that it does have something to do with cache size but this remains far from clear. and anyway, I've had a cache as large as 55 gbs on my laptop with no problems.

Oh and one other thing I've learned recently: Apple's Time Machine does not back up Bridge's caching system -- in fact it doesn't back up the User library's cache folder at all. Curious, eh?

Monday, September 5, 2011

"Error writing metadata to file"

I've talked about this problem early on in this blog since it started happening soon after the library went live some two years ago. And while it's not as catastrophic as a corrupted or damaged cache, it is still annoying and sometimes downright aggravating. I want to revisit the issue now because I've noticed a pattern of sorts and can't figure out what that means (if anything). Anyway, my hope is that someone reading this may have figured out what the problem is and how to fix it:


Here's what happens:
  1. I go into a folder of images in our library on the server and select a block of images to modify the metadata. 
  2. I make the metadata change(s), typically in either keywords or the description field.
  3. Bridge begins to modify the files -- note that my files are usually sorted by filename (lowest to highest number) but Bridge commences its "retagging" from highest to lowest. 
  4. It's at this point that Bridge appears to stall and soon brings up the above-noted error message on the first file; I click OK and the same thing happens with the next file; I click OK and then Bridge is off and running tagging all the remaining files just fine without any intervening error messages.
  5. I then go back to retag the files that Bridge refused to tag, one-by-one, and the updating works just fine -- no error message.
The frustrating aspect of all this if you are trying to modify a large block of images and the error message keeps repeating there is no way out of the cycle other than to force quit Bridge (no "escape" if you will).

Oh and by the way if I attempt to modify the metadata in the search results window instead of the actual folder, more often than not I'll get the error message far more frequently and across more images in the selected group.

The problem has occurred
  • with various file formats (but typically with JPG since those make up the bulk of our library)
  • with few if any additional programs open (no pattern noticed)
I've noticed this on two benchmark (fully indexed) Macs:
  • a desktop running Leopard 10.5.8 and 
  • a laptop running SNow Leopard 10.6.8
  • no, I haven't tried tested to see whether it occurs on a PC