* [f2fs-dev] [PATCH] f2fs: change write_hint for hot/warm nodes
[not found] <CGME20201103064039epcms2p30ecac0e7cefff0d50745f2e2e61ce38e@epcms2p3>
@ 2020-11-03 6:40 ` Daejun Park
2020-11-03 8:28 ` Chao Yu
[not found] ` <CGME20201103064039epcms2p30ecac0e7cefff0d50745f2e2e61ce38e@epcms2p8>
0 siblings, 2 replies; 3+ messages in thread
From: Daejun Park @ 2020-11-03 6:40 UTC (permalink / raw)
To: jaegeuk, chao
Cc: linux-f2fs-devel, linux-kernel, SEUNGUK SHIN, Sang-yoon Oh,
Sung-Jun Park, yongmyung lee, Daejun Park, Jinyoung CHOI,
Jieon Seol, Mankyu PARK, Jaemyung Lee, Keoseong Park
From 818a76a9aee5bf225565264274d211edb07bae7d Mon Sep 17 00:00:00 2001
From: Daejun Park <daejun7.park@samsung.com>
Date: Tue, 3 Nov 2020 15:30:26 +0900
In the fs-based mode of F2FS, the mapping of hot/warm node to
WRITE_LIFE_NOT_SET should be changed to WRITE_LIFE_SHORT.
As a result of analyzing the write pattern of f2fs using real workload,
hot/warm nodes have high update ratio close to hot data.[*]
However, F2FS passes write hints for hot/warm nodes as WRITE_LIFE_NOT_SET.
Furthermore, WRITE_LIFE_NOT_SET is a default value of write hint when it is
not provided from the file system.
In storage, write_hint is used to distinguish streams (e.g. NVMe).
So, the hot/warm node of F2FS is not distinguished from other write_hints,
which can make the wrong stream seperation.
* Liang, Yu, et al. "An empirical study of F2FS on mobile devices." 2017
IEEE 23rd International Conference on Embedded and Real-Time Computing
Systems and Applications (RTCSA).
Signed-off-by: Daejun Park <daejun7.park@samsung.com>
---
fs/f2fs/segment.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 1596502f7375..7b42bb10c6c3 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -3208,7 +3208,7 @@ enum rw_hint f2fs_io_type_to_rw_hint(struct f2fs_sb_info *sbi,
return WRITE_LIFE_EXTREME;
} else if (type == NODE) {
if (temp == WARM || temp == HOT)
- return WRITE_LIFE_NOT_SET;
+ return WRITE_LIFE_SHORT;
else if (temp == COLD)
return WRITE_LIFE_NONE;
} else if (type == META) {
--
2.25.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [f2fs-dev] [PATCH] f2fs: change write_hint for hot/warm nodes
2020-11-03 6:40 ` [f2fs-dev] [PATCH] f2fs: change write_hint for hot/warm nodes Daejun Park
@ 2020-11-03 8:28 ` Chao Yu
[not found] ` <CGME20201103064039epcms2p30ecac0e7cefff0d50745f2e2e61ce38e@epcms2p8>
1 sibling, 0 replies; 3+ messages in thread
From: Chao Yu @ 2020-11-03 8:28 UTC (permalink / raw)
To: daejun7.park, jaegeuk, chao
Cc: yongmyung lee, Jieon Seol, Sang-yoon Oh, Mankyu PARK,
Sung-Jun Park, linux-kernel, linux-f2fs-devel, Keoseong Park,
SEUNGUK SHIN, Jinyoung CHOI, Jaemyung Lee
On 2020/11/3 14:40, Daejun Park wrote:
>>From 818a76a9aee5bf225565264274d211edb07bae7d Mon Sep 17 00:00:00 2001
> From: Daejun Park <daejun7.park@samsung.com>
> Date: Tue, 3 Nov 2020 15:30:26 +0900
>
>
> In the fs-based mode of F2FS, the mapping of hot/warm node to
> WRITE_LIFE_NOT_SET should be changed to WRITE_LIFE_SHORT.
>
> As a result of analyzing the write pattern of f2fs using real workload,
> hot/warm nodes have high update ratio close to hot data.[*]
> However, F2FS passes write hints for hot/warm nodes as WRITE_LIFE_NOT_SET.
>
> Furthermore, WRITE_LIFE_NOT_SET is a default value of write hint when it is
> not provided from the file system.
> In storage, write_hint is used to distinguish streams (e.g. NVMe).
> So, the hot/warm node of F2FS is not distinguished from other write_hints,
> which can make the wrong stream seperation.
>
> * Liang, Yu, et al. "An empirical study of F2FS on mobile devices." 2017
> IEEE 23rd International Conference on Embedded and Real-Time Computing
> Systems and Applications (RTCSA).
Could you please update Documentation/filesystems/f2fs.rst as well?
Thanks,
>
> Signed-off-by: Daejun Park <daejun7.park@samsung.com>
> ---
> fs/f2fs/segment.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> index 1596502f7375..7b42bb10c6c3 100644
> --- a/fs/f2fs/segment.c
> +++ b/fs/f2fs/segment.c
> @@ -3208,7 +3208,7 @@ enum rw_hint f2fs_io_type_to_rw_hint(struct f2fs_sb_info *sbi,
> return WRITE_LIFE_EXTREME;
> } else if (type == NODE) {
> if (temp == WARM || temp == HOT)
> - return WRITE_LIFE_NOT_SET;
> + return WRITE_LIFE_SHORT;
> else if (temp == COLD)
> return WRITE_LIFE_NONE;
> } else if (type == META) {
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* RE: Re: [f2fs-dev] [PATCH] f2fs: change write_hint for hot/warm nodes
[not found] ` <CGME20201103064039epcms2p30ecac0e7cefff0d50745f2e2e61ce38e@epcms2p8>
@ 2020-11-03 8:33 ` Daejun Park
0 siblings, 0 replies; 3+ messages in thread
From: Daejun Park @ 2020-11-03 8:33 UTC (permalink / raw)
To: Chao Yu, Daejun Park, jaegeuk, chao
Cc: yongmyung lee, Jieon Seol, Sang-yoon Oh, Mankyu PARK,
Sung-Jun Park, linux-kernel, linux-f2fs-devel, Keoseong Park,
SEUNGUK SHIN, Jinyoung CHOI, Jaemyung Lee
>>>From 818a76a9aee5bf225565264274d211edb07bae7d Mon Sep 17 00:00:00 2001
>> From: Daejun Park <daejun7.park@samsung.com>
>> Date: Tue, 3 Nov 2020 15:30:26 +0900
>>
>>
>> In the fs-based mode of F2FS, the mapping of hot/warm node to
>> WRITE_LIFE_NOT_SET should be changed to WRITE_LIFE_SHORT.
>>
>> As a result of analyzing the write pattern of f2fs using real workload,
>> hot/warm nodes have high update ratio close to hot data.[*]
>> However, F2FS passes write hints for hot/warm nodes as WRITE_LIFE_NOT_SET.
>>
>> Furthermore, WRITE_LIFE_NOT_SET is a default value of write hint when it is
>> not provided from the file system.
>> In storage, write_hint is used to distinguish streams (e.g. NVMe).
>> So, the hot/warm node of F2FS is not distinguished from other write_hints,
>> which can make the wrong stream seperation.
>>
>> * Liang, Yu, et al. "An empirical study of F2FS on mobile devices." 2017
>> IEEE 23rd International Conference on Embedded and Real-Time Computing
>> Systems and Applications (RTCSA).
>
>Could you please update Documentation/filesystems/f2fs.rst as well?
>
>Thanks,
Sure, I will make it next patch.
Thanks
>>
>> Signed-off-by: Daejun Park <daejun7.park@samsung.com>
>> ---
>> fs/f2fs/segment.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
>> index 1596502f7375..7b42bb10c6c3 100644
>> --- a/fs/f2fs/segment.c
>> +++ b/fs/f2fs/segment.c
>> @@ -3208,7 +3208,7 @@ enum rw_hint f2fs_io_type_to_rw_hint(struct f2fs_sb_info *sbi,
>> return WRITE_LIFE_EXTREME;
>> } else if (type == NODE) {
>> if (temp == WARM || temp == HOT)
>> - return WRITE_LIFE_NOT_SET;
>> + return WRITE_LIFE_SHORT;
>> else if (temp == COLD)
>> return WRITE_LIFE_NONE;
>> } else if (type == META) {
>>
>
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2020-11-03 8:33 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
[not found] <CGME20201103064039epcms2p30ecac0e7cefff0d50745f2e2e61ce38e@epcms2p3>
2020-11-03 6:40 ` [f2fs-dev] [PATCH] f2fs: change write_hint for hot/warm nodes Daejun Park
2020-11-03 8:28 ` Chao Yu
[not found] ` <CGME20201103064039epcms2p30ecac0e7cefff0d50745f2e2e61ce38e@epcms2p8>
2020-11-03 8:33 ` Daejun Park
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).