C1 wears out SSD by writing masks when just browsing!
When browsing through the library and browser (viewing thumbnails, not even the viewer is opened), C1 permanently reads and writes existing masks to disk.
Not only that each mask is written once, as I browse through different collections, the same .comask file is written several times.
I have 4000 images in catalog, roughly 350 comask files totalling 1.3 GB, when I browse through my most beloved collections roughly 210 comask files (totalling 850 MB) are written with a new timestamp, I can see this in windows explorer, ordered by last modified date.
It seems not only some header is written but the total file, as you can see the massive amount of data written with the sysinternals tool "process monitor", menu > tools > file summary:
http://i66.tinypic.com/14nl5w3.jpg
C1 reads 2.5 GB and writes 1.1 GB though I only browsed (scrolled) for about 10 minutes in C1 (interrupted by some beachball mousepointer and unresponsiveness of C1).
Not only that is wears out disk drives I assume it also decreases performance, and it is a totally unnecessary action! The only time when .comask files should be updated is when modifying layers or brushing / erasing masks.
What is your experience and opinion?
Cheers
BeO
P.S. This is present in v8 and v9 including 9.1.2.
Not only that each mask is written once, as I browse through different collections, the same .comask file is written several times.
I have 4000 images in catalog, roughly 350 comask files totalling 1.3 GB, when I browse through my most beloved collections roughly 210 comask files (totalling 850 MB) are written with a new timestamp, I can see this in windows explorer, ordered by last modified date.
It seems not only some header is written but the total file, as you can see the massive amount of data written with the sysinternals tool "process monitor", menu > tools > file summary:
http://i66.tinypic.com/14nl5w3.jpg
C1 reads 2.5 GB and writes 1.1 GB though I only browsed (scrolled) for about 10 minutes in C1 (interrupted by some beachball mousepointer and unresponsiveness of C1).
Not only that is wears out disk drives I assume it also decreases performance, and it is a totally unnecessary action! The only time when .comask files should be updated is when modifying layers or brushing / erasing masks.
What is your experience and opinion?
Cheers
BeO
P.S. This is present in v8 and v9 including 9.1.2.
0
-
Hi BeO,
We tried to contact you on your email, but our mails from various accounts are being bounced by your mail-server:
xxxxxx@xxxxxx
host xxxxxxx [xxxxxx]
SMTP error from remote mail server after MAIL FROM:<xxxxx@xxxxxxxxx.com> SIZE=xxxxxxxx:
550-Requested action not taken: mailbox unavailable
550-Reject due to SPF policy.
550-The originating IP of the message is not permitted by the domain owner.
550 For explanation visit xxxxxxx0 -
Thanks Christian.
I just checked my box and 2 mails from your colleague M. came through this morning.
Cheers
BeO0 -
Good to hear! 0 -
That's great that you are emailing BeO, but can you discuss it here too please. Sounds kinda important. 0 -
Hi,
An additional and related remark...
After reading meanwhile's comments, I launched C1 and also the Process Monitor. I put the Process Monitor to the foreground, so C1 was not doing anything (apparently).
I just observed what it was doing in the background and I could see that it is constantly re-reading the display ICC profiles (I have a dual monitor configuration). It permanently reads both files again and again even if the user doesn't interact with C1.
I noticed that both ICC profiles are read by using the CreateFileMapping API, that is, through a memory-mapped file. This is a good idea since this is typically the kind of data that should be cached in memory and used via a memory mapped file in order to limit disk usage. So why on earth constantly open and close these files and re-create the memory mapped file each time? Just keep it open for the whole session.0 -
[quote="meanwhile" wrote:
That's great that you are emailing BeO, but can you discuss it here too please. Sounds kinda important.
I can't say more right now but the fact that P1 emailed me indicates that P1 might share your assessment 😊0 -
Would someone from Phase One please provide some further information on this topic?
If the issue is as concerning as it seems to be then I think more users than just the OP should be a part of the discussion.
If, on the other hand, it's something that we all shouldn't be concerned about then I think that we should be told that as well.
Thanks.0 -
Hi Philbond,
Regarding SSD wear-out, everybody who has a crystal ball can give you a definite answer, but there are only a few out there and neither I nor P1 staff has one 😊
SSDs wear out with write operations due to the physics of SSDs. How long the "life-time" of a specific SSD under a specific usage actually is is something probably no one can answer correctly, but there are tools available which give you an indication about the "health" state of your drive. There is a lot of information in the internet.
For this specific issue, the question is how much data is written to your specific disk when you work with C1, and this depends on a lot of factors, mainly how many images with masks you have, how big they are and whether you browse through a collection of images with masks and how often.
In my example I have a 4k catalog and browsed through a lot of images with masks which is something I do not do every day as I am not a pro (I haven't measured written bytes on a "full working day", though). C1 wrote roughly 1 GB unnessessarily in my test, and if it does this every single day in a month this equates to roughly one 32 GB SD card full of images copied from the camera to disk.
Hence I think you should be concerned somewhat but not too much.
I will be happy to update C1 as soon as a fix is available though. As a beta tester I am not entitled to comment on this, and P1 rarely comments in this user forum on upcoming things either, so just stay tuned.
Cheers
BeO0
Post is closed for comments.
Comments
8 comments