Weird behavior writing to SSD on 2013 MacBook
Dimitry Andric
dim at FreeBSD.org
Thu Feb 5 07:57:10 UTC 2015
If you let bsdtar continue, and press control-T a few times, does the
user time (u) increase at all? Does it ever go any further, if you let
it run for a very long time?
I believe a problem may have been introduced by r277922, leading to
filesystem hangs in some scenarios. It looks like this commit is also
in dumbbell's github fork:
https://github.com/dumbbell/freebsd/commit/83723416a6bb8695d60c6573722a81086899f521
-Dimitry
> On 05 Feb 2015, at 08:32, Lundberg, Johannes <johannes at brilliantservice.co.jp> wrote:
>
> The release is the latest snapshot memstick image and kernel is from
> https://github.com/dumbbell/freebsd/tree/kms-drm-update-38.
>
> I think this problem existed before I changed kernel.
>
> I haven't changed any settings on the filesystem so I assume TRIM is off.
>
> When stopped "bsdtar" is in state "flswai". What does this mean?
>
> / is 13 GB and 80% full.
>
>
> --
> Johannes Lundberg
> BRILLIANTSERVICE CO., LTD.
>
> On Thu, Feb 5, 2015 at 4:21 PM, Allan Jude <allanjude at freebsd.org> wrote:
>
>> On 2015-02-04 20:40, Lundberg, Johannes wrote:
>>> By the way,
>>>
>>> For the second test I first ran portsnap extract without removing the old
>>> /usr/ports folder and it ran through quickly without any halts..
>>>
>>> --
>>> Johannes Lundberg
>>> BRILLIANTSERVICE CO., LTD.
>>>
>>> On Thu, Feb 5, 2015 at 10:38 AM, Lundberg, Johannes <
>>> johannes at brilliantservice.co.jp> wrote:
>>>
>>>> I deleted /usr/ports and did a new portsnap extract
>>>>
>>>> portsnap stopped at /usr/ports/editors/teco
>>>>
>>>> that folder is empty and the previous folder (editors/tea) is populated
>>>> with files.
>>>>
>>>> portsnap stopped for about 2-3 minutes and during the whole time gstat
>>>> showed values like this: (disc io load was constantly fluctuating around
>>>> 200 MB/s, not static)
>>>>
>>>> dT: 1.002s w: 1.000s
>>>> L(q) ops/s r/s kBps ms/r w/s kBps ms/w %busy Name
>>>> 1240 43523 0 0 0.0 43523 220158 24.1 99.5| ada0
>>>> 0 0 0 0 0.0 0 0 0.0 0.0| ada0p1
>>>> 0 0 0 0 0.0 0 0 0.0 0.0| ada0p2
>>>> 0 0 0 0 0.0 0 0 0.0 0.0| ada0p3
>>>> 0 0 0 0 0.0 0 0 0.0 0.0| ada0p4
>>>> 0 0 0 0 0.0 0 0 0.0 0.0| ada0p5
>>>> 0 0 0 0 0.0 0 0 0.0 0.0| ada0p6
>>>> 1240 43523 0 0 0.0 43523 220158 24.1 99.5| ada0p7
>>>> 0 0 0 0 0.0 0 0 0.0 0.0| ada0p8
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gpt/EFI%20System%20Partition
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gptid/ca33c17c-0ef4-4d9b-b2bb-cb37a907504b
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>> msdosfs/EFI
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>> gpt/Untitled
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gptid/319461e8-0310-47d5-b4d1-6ba5a92cf9a9
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gpt/Recovery%20HD
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gptid/cb9530b7-8872-46d0-b36c-fca667b4e541
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gptid/6ac11466-21c5-4420-85bc-eb1c3c7fa616
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gptid/0047cc59-6b75-4508-98d0-842beafd3164
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> gptid/ddebb168-ac18-11e4-8f9e-283737012e32
>>>> 0 0 0 0 0.0 0 0 0.0 0.0|
>>>> msdosfs/NO_NAME
>>>>
>>>>
>>>> That is, 100% busy and 200 MB/s...
>>>>
>>>> top shows
>>>>
>>>> last pid: 13709; load averages: 1.18, 0.98,
>>>> 0.58
>>>> up 0+00:28:36 10:35:38
>>>> 27 processes: 1 running, 26 sleeping
>>>> CPU: 0.0% user, 0.0% nice, 12.3% system, 11.1% interrupt, 76.6% idle
>>>> Mem: 25M Active, 651M Inact, 587M Wired, 30M Cache, 411M Buf, 2566M Free
>>>> Swap: 706M Total, 706M Free
>>>>
>>>>
>>>> I have used FreeBSD with SSD plenty and never seen this behavior before.
>>>>
>>>> --
>>>> Johannes Lundberg
>>>> BRILLIANTSERVICE CO., LTD.
>>>>
>>>> On Thu, Feb 5, 2015 at 10:14 AM, Allan Jude <allanjude at freebsd.org>
>> wrote:
>>>>
>>>>> On 2015-02-04 19:29, Lundberg, Johannes wrote:
>>>>>> Hi
>>>>>>
>>>>>> I'm thought I was gonna do some test runs with HEAD on a 2013 Macbook
>>>>> Air
>>>>>> and noticed some weird behavior regarding disk I/O.
>>>>>>
>>>>>> This happens both when doing portsnap extract and clone from git
>>>>> repository.
>>>>>>
>>>>>> For example portsnap extract, the extraction process (the output of
>> it)
>>>>>> suddenly stops, for seconds or maybe even minutes, quite many times
>>>>> during
>>>>>> the whole extraction process.
>>>>>> iostat reports ~200 MB/s on ada0 the whole time during freeze.
>>>>>>
>>>>>>
>>>>>> pciconf:
>>>>>>
>>>>>> ahci0 at pci0:4:0:0: class=0x010601 card=0x91831b4b chip=0x91831b4b
>>>>>> rev=0x14 hdr=0x00
>>>>>> vendor = 'Marvell Technology Group Ltd.'
>>>>>> class = mass storage
>>>>>> subclass = SATA
>>>>>>
>>>>>>
>>>>>> dmesg (relevant lines?):
>>>>>>
>>>>>> ahci0: <Marvell 88SS9183 AHCI SATA controller> port
>>>>>> 0x1028-0x102f,0x1034-0x1037,0x1020-0x1027,0x1030-0x1033,0x1000-0x101f
>>>>> mem
>>>>>> 0xb0700000-0xb07001ff at device 0.0 on pci4
>>>>>> ahci0: AHCI v1.00 with 1 6Gbps ports, Port Multiplier not supported
>>>>>> ahcich0: <AHCI channel> at channel 0 on ahci0
>>>>>>
>>>>>> ada0 at ahcich0 bus 0 scbus0 target 0 lun 0
>>>>>> ada0: <APPLE SSD SD0128F A222821> ATA-8 SATA 3.x device
>>>>>> ada0: Serial Number 1325A5401681 \^T\^T\^T\^T\^T\^T\^T
>>>>>> ada0: 600.000MB/s transfers (SATA 3.x, UDMA6, PIO 512bytes)
>>>>>> ada0: Command Queueing enabled
>>>>>> ada0: 115712MB (236978176 512 byte sectors: 16H 63S/T 16383C)
>>>>>> ada0: Previously was known as ad4
>>>>>>
>>>>>> GEOM: ada0: enabling Boot Camp
>>>>>> GEOM: diskid/DISK-1325A5401681%20%14%14%14%14%14%14%14: enabling Boot
>>>>> Camp
>>>>>>
>>>>>>
>>>>>> gpart:
>>>>>>
>>>>>> => 34 236978109 ada0 GPT (113G)
>>>>>> 34 6 - free - (3.0K)
>>>>>> 40 409600 1 efi (200M)
>>>>>> 409640 174519128 2 apple-hfs (83G)
>>>>>> 174928768 1269536 3 apple-boot (620M)
>>>>>> 176198304 1376 - free - (688K)
>>>>>> 176199680 29782016 4 linux-data (14G)
>>>>>> 205981696 2097152 5 linux-swap (1.0G)
>>>>>> 208078848 1600 6 efi (800K)
>>>>>> 208080448 27261368 7 freebsd-ufs (13G)
>>>>>> 235341816 1445888 8 freebsd-swap (706M)
>>>>>> 236787704 190439 - free - (93M)
>>>>>>
>>>>>>
>>>>>> One other weird thing is that FreeBSD does not show up in the refind
>>>>> boot
>>>>>> menu by default, only OSX and Linux. I have to press ESC once to
>> reload
>>>>> for
>>>>>> FreeBSD boot option to show up.. Any clues? Is my partition
>>>>> configuration
>>>>>> wrong in some way?
>>>>>>
>>>>>> Thanks!
>>>>>> --
>>>>>> Johannes Lundberg
>>>>>> BRILLIANTSERVICE CO., LTD.
>>>>>>
>>>>>
>>>>> For the disk io bit, try running 'gstat' instead of iostat, and see
>> what
>>>>> it says.
>>>>>
>>>>> --
>>>>> Allan Jude
>>>>>
>>>>>
>>>>
>>>
>>
>> Is the disk nearly full? very random guess, but maybe it is the SSD
>> running its garbage collection when it runs out of space. Do you have
>> TRIM enabled?
>>
>> --
>> Allan Jude
>>
>>
>
> --
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
> 秘密保持について:この電子メールは、名宛人に送信したものであり、秘匿特権の対象となる情報を含んでいます。
> もし、名宛人以外の方が受信された場合、このメールの破棄、およびこのメールに関する一切の開示、
> 複写、配布、その他の利用、または記載内容に基づくいかなる行動もされないようお願い申し上げます。
> ---
> CONFIDENTIALITY NOTE: The information in this email is confidential
> and intended solely for the addressee.
> Disclosure, copying, distribution or any other action of use of this
> email by person other than intended recipient, is prohibited.
> If you are not the intended recipient and have received this email in
> error, please destroy the original message.
> _______________________________________________
> freebsd-current at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscribe at freebsd.org"
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 194 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.freebsd.org/pipermail/freebsd-current/attachments/20150205/b96041de/attachment.sig>
More information about the freebsd-current
mailing list