[Partclone-user] partclone feature requests
Brought to you by:
thomas_tsai
|
From: Naoyuki T. <nao...@gm...> - 2019-07-17 16:41:30
|
Hello, First, I have to thank for partclone, as it's vital component of my volunteer work. Well, I have a couple of feature requests. Not sure how well sourceforge's tickets are cared for so I'm trying to send a mail. #1 - Flush out the cache while restoring When I am restoring to USB flash drive, there is a long long long pause between "99%" and "100%". I am looking at the code and I can guess why. "fs_close()" is called after restoring and 100% shows up after fs_close() is done. The machine with a lot of memory these days can hold a lot of dirty pages, and fs_close() forces the dirty cache to be flushed to USB flash drive, and this writing takes longer than actual restoring process. For example, I restore 2.7GB disk image to a USB flash drive, last progress before 100 shows up at 44 seconds from the start. 100% progress shows up after 225 seconds. That means, partclone is running but there is about 180 seconds of silence between <100 and 100% progress. If there is a cache flush after each block is written to the device, the progress would be far more accurate. Looking at the code, I can guess it's probably exfat_flush_nodes or exfat_flush does the job but I obviously don't have the knowledge. If you want me to try something, I'm happy to. #2 - Progress output format. I'm parsing the output of partclone using Python. It would be great if the output is more machine readable. Thanks and have a nice day. -- Tai P.S. Volunteer work is at https://github.com/ntai/wce-triage-v2 https://github.com/ntai/wce-triage-v2/blob/master/wce_triage/bin/restore_volume.py |