Red Hat Bugzilla – Bug 488123
Booting with any entries in /etc/crypttab adds 10 seconds to boot
Last modified: 2009-06-30 16:19:10 EDT
Created attachment 333781 [details]
bootchart booting with entry in /etc/crypttab
Description of problem:
Not sure this is a plymouth problem, so "sorry in advance" if I've misdirected this.
I have the following setup: a laptop (thinkpad X200) with 2 encrypted partitions. One is '/' (/dev/mapper/VolGroup00-LogVol00, created/setup during installation), the second (/dev/sda2, created after installation).
Both are ext3, and both can be decrypted/mounted using the same passphrase.
Plymouth and rc.sysinit work fine to prompt for the passphrase one time and mount both.
Here is the entry in /etc/crypttab for the "second" partition:
luks-49955702-1c8b-4f5b-9ab3-0ce4ee2873e9 UUID=49955702-1c8b-4f5b-9ab3-0ce4ee2873e9 none
Booting with just this asks for the passphrase once, and displays the second partition mounted (with an nice "unlocked icon" on the desktop) as /media/extra (label for this partition is "extra").
I attach 2 bootcharts: the first booting as described above, the second booting with the entry in /etc/crypttab commented out.
The second boot seems to take 10 seconds less.
The first bootchart shows an "extra" plymouth and a cryptsetup process (from about 28 seconds to about 45 seconds). Also see an "extra" pidof process running a bit less.
Also notice that my bootcharts all "start" about 15 seconds from "T==0". Not sure if this is related....
Version-Release number of selected component (if applicable):
Steps to Reproduce:
Created attachment 333782 [details]
bootchart booting with commented out /etc/crypttab
Bootchart when entry is commented out in /etc/crypttab.
Some additional info:
Booting "without crypttab" and manually running cryptsetup luksOpen shows a much faster run time:
echo -n "passphrase" | time cryptsetup --key-file=- luksOpen /dev/sda2 extra
key slot 0 unlocked.
0.47user 0.12system 0:01.94elapsed 30%CPU (0avgtext+0avgdata 0maxresident)k
127inputs+0outputs (0major+21874minor)pagefaults 0swaps
So this runs in less than 2 seconds.
This bug appears to have been reported against 'rawhide' during the Fedora 11 development cycle.
Changing version to '11'.
More information and reason for this action is here:
Sorry, i can`t see a bug here. Everything works.
Thanks for your contribution to fedora.
Fedora Bugzappers volunteer triage team