All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Dolev Raviv" <draviv@codeaurora.org>
To: "'Richard Weinberger'" <richard.weinberger@gmail.com>
Cc: <linux-mtd@lists.infradead.org>, <linux-ext4@vger.kernel.org>,
	"'Tanya Brokhman'" <tlinder@codeaurora.org>
Subject: RE: planning general storage capacity for y fs
Date: Mon, 27 Apr 2015 10:55:51 +0300	[thread overview]
Message-ID: <007201d080bf$96a0e4e0$c3e2aea0$@codeaurora.org> (raw)
In-Reply-To: <CAFLxGvyJDyN2-WJUkJ9wz5cCDvNjrUsbdpHrN4DH7S1gQoF2Yw@mail.gmail.com>

>On Sun, Apr 26, 2015 at 1:14 PM, Dolev Raviv <draviv@codeaurora.org> wrote:
>> Hi,
>> I'm looking for guidelines for planning storage capacity. I understand 
>> it strongly depended on the usage type.
>> I want to know at what point storage fullness is effecting performance 
>> in a standard read/write partition. Do different File Systems 
>> (UBIFS/EXT4) have different full-free ratio?
>> What about read only fs? Can I plan less free space in such cases?
>>
>> I'll appreciate any input on this, for UBIFS specific and fs in general.
>
>Not sure if I got your question.
>You want to know how filesystems in general behave when they run out of free space?
>The general answer is that they need more effort to find free space.
>
>In case of UBIFS you also have to think of the garbage collector.
>See http://www.linux-mtd.infradead.org/doc/ubifs.html
>At the end of the day you'll have to run benchmarks on your own to find out how a specific filesystem behaves on your workload...
>

Thanks Richard!
Let me rephrase the question: In the past I knew there was a rule of thumb, 'leave free 30% of the storage space'. Nowadays I couldn't find any reference to this.
I was wondering, is there a known point in UBIFS (or ext4), where leaving less free storage space, that performance is dropping? Maybe a ratio of free-occupied is not the right way to look at it, but to leave a certain size free (e.g. 50MB)?

Thanks,
Dolev
-- 
Qualcomm Israel, on behalf of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project


WARNING: multiple messages have this Message-ID (diff)
From: "Dolev Raviv" <draviv@codeaurora.org>
To: "'Richard Weinberger'" <richard.weinberger@gmail.com>
Cc: linux-ext4@vger.kernel.org, linux-mtd@lists.infradead.org,
	'Tanya Brokhman' <tlinder@codeaurora.org>
Subject: RE: planning general storage capacity for y fs
Date: Mon, 27 Apr 2015 10:55:51 +0300	[thread overview]
Message-ID: <007201d080bf$96a0e4e0$c3e2aea0$@codeaurora.org> (raw)
In-Reply-To: <CAFLxGvyJDyN2-WJUkJ9wz5cCDvNjrUsbdpHrN4DH7S1gQoF2Yw@mail.gmail.com>

>On Sun, Apr 26, 2015 at 1:14 PM, Dolev Raviv <draviv@codeaurora.org> wrote:
>> Hi,
>> I'm looking for guidelines for planning storage capacity. I understand 
>> it strongly depended on the usage type.
>> I want to know at what point storage fullness is effecting performance 
>> in a standard read/write partition. Do different File Systems 
>> (UBIFS/EXT4) have different full-free ratio?
>> What about read only fs? Can I plan less free space in such cases?
>>
>> I'll appreciate any input on this, for UBIFS specific and fs in general.
>
>Not sure if I got your question.
>You want to know how filesystems in general behave when they run out of free space?
>The general answer is that they need more effort to find free space.
>
>In case of UBIFS you also have to think of the garbage collector.
>See http://www.linux-mtd.infradead.org/doc/ubifs.html
>At the end of the day you'll have to run benchmarks on your own to find out how a specific filesystem behaves on your workload...
>

Thanks Richard!
Let me rephrase the question: In the past I knew there was a rule of thumb, 'leave free 30% of the storage space'. Nowadays I couldn't find any reference to this.
I was wondering, is there a known point in UBIFS (or ext4), where leaving less free storage space, that performance is dropping? Maybe a ratio of free-occupied is not the right way to look at it, but to leave a certain size free (e.g. 50MB)?

Thanks,
Dolev
-- 
Qualcomm Israel, on behalf of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project

  reply	other threads:[~2015-04-27  7:55 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-04-26 11:14 planning general storage capacity for y fs Dolev Raviv
2015-04-26 20:45 ` Richard Weinberger
2015-04-26 20:45   ` Richard Weinberger
2015-04-27  7:55   ` Dolev Raviv [this message]
2015-04-27  7:55     ` Dolev Raviv
2015-04-27  8:06     ` Richard Weinberger
2015-04-27  8:06       ` Richard Weinberger

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='007201d080bf$96a0e4e0$c3e2aea0$@codeaurora.org' \
    --to=draviv@codeaurora.org \
    --cc=linux-ext4@vger.kernel.org \
    --cc=linux-mtd@lists.infradead.org \
    --cc=richard.weinberger@gmail.com \
    --cc=tlinder@codeaurora.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.