All of lore.kernel.org
 help / color / mirror / Atom feed
From: Zachary Palmer <zep_bcache-J5qI5MFTcs8@public.gmane.org>
To: linux-bcache-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: bcache on a laptop
Date: Thu, 19 Sep 2013 19:10:03 -0400	[thread overview]
Message-ID: <523B844B.8070202@bahj.com> (raw)

Gents,

I'm still trying to get a comfy bcache installation on a laptop.  To 
summarize thus far,

     1. I have a Debian system that boots from bcache
     2. No version of 3.10 that I have tested (either from the Linux git 
or from the bcache-stable branch of the bcache git) will allow the 
laptop to sleep or hibernate properly.
     3. Every version of 3.11 that I have tested results in severe data 
corruption.  (I wound up having to reinstall my system as a result.)  
Using 3.11.0 built from the 3.11.y branch at kernel.org, the corruption 
was subtle at first; eventually programs started segfaulting.  Running 
"sync ; echo 3 > /proc/sys/vm/drop_caches" would resolve the corruption 
issue (I'm in writethrough mode), but the root was destroyed when 
corruption occurred in the middle of a nightly upgrade.
     4. When corruption isn't an issue (e.g. in a 3.10 kernel), I'm 
getting between a 50% and 80% cache hit ratio (again in writethrough, 
not writeback).  I'm pretty happy about that.  :)
     5. Creating a large number of small files seems to be quite a lot 
slower than it was without bcache.  (For instance, installing xfce4 and 
its dependencies into a fresh chroot takes about two hours.)  I've 
noticed that Debian packages in particular take longer; kernel header 
packages often take close to half an hour by themselves.  (Can anyone 
posit a reason for this?)

I'll be trying the 3.10.12 bcache-stable shortly, though I expect that 
it won't make a huge difference in the hibernation problem.  I just 
wanted to record this information in case someone else decides to give 
this a go.

Cheers,

Zach

                 reply	other threads:[~2013-09-19 23:10 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=523B844B.8070202@bahj.com \
    --to=zep_bcache-j5qi5mftcs8@public.gmane.org \
    --cc=linux-bcache-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.