Looking for the paradigm » ssd http://www.evanhoffman.com/evan So I can pass it off Thu, 26 Jan 2012 22:17:15 +0000 en hourly 1 http://wordpress.org/?v=3.3.1 Macbook Pro locks up with SSD installed. http://www.evanhoffman.com/evan/2011/08/23/macbook-pro-locks-up-with-ssd-installed/?utm_source=rss&utm_medium=rss&utm_campaign=macbook-pro-locks-up-with-ssd-installed http://www.evanhoffman.com/evan/2011/08/23/macbook-pro-locks-up-with-ssd-installed/#comments Tue, 23 Aug 2011 16:04:54 +0000 Evan Hoffman http://www.evanhoffman.com/evan/?p=1563

A few weeks ago I switched from my trusty old HP nc8430 to a Macbook Pro (MC118LL/A) that was left spare when another employee left. I mostly enjoyed using Linux but I was tired of dealing with weird quirks like having X lock up, essentially forcing me to do a hard reboot.

To transition, I copied my documents from Linux to Mac, then turned off the Linux laptop. Surprisingly I found I didn’t need to turn Linux back on at all.

Last week, I decided to put the final nail in Linux’s coffin by taking the SSD (Corsair CMFSSD-128GBG2D)out of it and putting it in my Macbook. The Macbook was pretty fast (Core 2 Duo @ 2.5 GHz) but some things were noticeably slower on its 7200RPM disk than on Linux with an SSD, especially running Windows VMs.

I booted Linux to Knoppix and zeroed out the disk, then removed it. I backed my Mac up with Time Machine, shut it down, then undid the ~12 tiny screws, removed the bottom plate of the Macbook and popped the SSD in. I booted from the Mac DVD, restored from Time Machine and went home (it took ~4 hours to restore).

When I got in the next day, the restore was complete, though I had to click “Restart” to finish, which was annoying. Everything worked fine, and I was pretty impressed. The machine was kind of sluggish due to Spotlight indexing but once that was done I was pretty amazed at the transformation. Every app opened in under 1 second. Windows VMs were super snappy. Things were going well.

But then I started noticing periods of extended hanging. In the middle of some task, I’d get the beachball and the whole computer would become unresponsive (cursor would spin & move around but I couldn’t click anything). This would last about 30-60 seconds. I assumed it was some behind-the-scenes optimization, or some residual spotlight indexing.

Unfortunately, it hasn’t gone away. Earlier today I copied a 3 GB zip file from our file server to my laptop and it beachballed me on and off (about 60-70% of the time) for about 15 minutes as it copied. What’s odd is that the transfer speeds were pretty good, it appeared to be my computer itself that was bottlenecking it. After the download completed, I attempted to unzip it and was beachballed again. I checked Activity Monitor and it was peaking at 30 MB/s, but had extended periods of zeroes. I ran iostat and got basically the same information:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89

EvanMBP:~ root# iostat -Kw 3
          disk0           disk1       cpu     load average
    KB/t tps  MB/s     KB/t tps  MB/s  us sy id   1m   5m   15m
    9.33   3  0.03     0.00   0  0.00   9  4 87  0.25 0.28 0.32
    0.00   0  0.00     0.00   0  0.00  13  5 82  0.23 0.27 0.32
   20.00   0  0.01     0.00   0  0.00  15  5 80  0.23 0.27 0.32
   36.00   1  0.02     0.00   0  0.00  16  6 77  0.21 0.27 0.32
    0.00   0  0.00     0.00   0  0.00  12  5 83  0.21 0.27 0.32
   20.02  60  1.18     0.00   0  0.00  17 10 74  0.19 0.26 0.32
   24.60 363  8.72     0.00   0  0.00  18 13 69  0.17 0.26 0.31
   25.40 307  7.60     0.00   0  0.00  14 11 75  0.17 0.26 0.31
   21.95 426  9.14     0.00   0  0.00  15 12 73  0.32 0.29 0.32
   82.50 352 28.35     0.00   0  0.00  17 11 73  0.32 0.29 0.32
  809.70  84 66.41     0.00   0  0.00  22  9 69  0.29 0.28 0.32
    0.00   0  0.00     0.00   0  0.00   9  5 86  0.27 0.28 0.32
    0.00   0  0.00     8.89  11  0.09   9  5 86  0.27 0.28 0.32
    0.00   0  0.00     0.00   0  0.00   9  5 87  0.33 0.29 0.32
    0.00   0  0.00     0.00   0  0.00  12  7 81  0.33 0.29 0.32
    0.00   0  0.00     0.00   0  0.00  14  8 78  0.38 0.30 0.33
    0.00   0  0.00     0.00   0  0.00  11  6 83  0.35 0.29 0.33
    0.00   0  0.00     0.00   0  0.00  10  6 84  0.35 0.29 0.33
    0.00   0  0.00     0.00   0  0.00  10  5 84  0.32 0.29 0.32
    0.00   0  0.00     0.00   0  0.00  11  6 84  0.32 0.29 0.32
          disk0           disk1       cpu     load average
    KB/t tps  MB/s     KB/t tps  MB/s  us sy id   1m   5m   15m
    0.00   0  0.00     0.00   0  0.00  10  5 84  0.30 0.28 0.32
    0.00   0  0.00     0.00   0  0.00  11  6 84  0.27 0.28 0.32
    0.00   0  0.00     0.00   0  0.00  11  5 84  0.27 0.28 0.32
    0.00   0  0.00     0.00   0  0.00   9  5 85  0.25 0.27 0.32
    0.00   0  0.00     0.00   0  0.00  10  5 85  0.25 0.27 0.32
    0.00   0  0.00     0.00   0  0.00  10  5 85  0.31 0.29 0.32
    0.00   0  0.00     0.00   0  0.00  10  5 85  0.45 0.31 0.33
    0.00   0  0.00     0.00   0  0.00  10  5 85  0.45 0.31 0.33
    0.00   0  0.00     0.00   0  0.00  10  5 85  0.41 0.31 0.33
    0.00   0  0.00     0.00   0  0.00  10  5 85  0.41 0.31 0.33
  384.49  10  3.85    10.00   0  0.00  12  6 82  0.32 0.29 0.32
  291.73 126 35.78     0.00   0  0.00  23 13 64  0.38 0.31 0.33
  236.65 338 78.20     0.00   0  0.00  30 17 53  0.34 0.30 0.33
  397.61  21  8.02     0.00   0  0.00  15  8 77  0.34 0.30 0.33
    0.00   0  0.00     0.00   0  0.00  11  6 83  0.32 0.30 0.32
    0.00   0  0.00     0.00   0  0.00  12  7 81  0.32 0.30 0.32
    0.00   0  0.00     0.00   0  0.00  12  6 82  0.29 0.29 0.32
    0.00   0  0.00     0.00   0  0.00  11  6 83  0.35 0.30 0.33
    0.00   0  0.00     0.00   0  0.00  13  6 80  0.35 0.30 0.33
    0.00   0  0.00     0.00   0  0.00  11  6 83  0.32 0.30 0.32
          disk0           disk1       cpu     load average
    KB/t tps  MB/s     KB/t tps  MB/s  us sy id   1m   5m   15m
    0.00   0  0.00     0.00   0  0.00  13  8 78  0.32 0.30 0.32
    0.00   0  0.00     0.00   0  0.00  11  7 82  0.29 0.29 0.32
    0.00   0  0.00     0.00   0  0.00  10  5 86  0.35 0.30 0.33
  148.94 124 18.03     0.00   0  0.00  18  8 73  0.35 0.30 0.33
  267.73 121 31.72     0.00   0  0.00  17  8 76  0.32 0.30 0.32
  355.54 162 56.35     0.00   0  0.00  22  8 69  0.32 0.30 0.32
  738.07  38 27.38     0.00   0  0.00  16  6 78  0.30 0.29 0.32
  512.42  67 33.52     0.00   0  0.00  20  7 73  0.27 0.29 0.32
  835.74  61 50.05     0.00   0  0.00  19  7 73  0.27 0.29 0.32
  536.83  69 36.17     0.00   0  0.00  17  6 76  0.49 0.33 0.34
  543.89  83 43.90     0.00   0  0.00  20  8 72  0.49 0.33 0.34
  720.70  59 41.74     0.00   0  0.00  18  7 76  0.45 0.33 0.33
  541.23 124 65.70     0.00   0  0.00  22  9 70  0.41 0.32 0.33
  260.54 210 53.37     0.00   0  0.00  22  9 70  0.41 0.32 0.33
  806.93  73 57.78     0.00   0  0.00  20  8 72  0.46 0.33 0.34
  874.98  43 37.02     0.00   0  0.00  13  7 80  0.46 0.33 0.34
    0.00   0  0.00     0.00   0  0.00  11  4 85  0.42 0.33 0.33
    0.00   0  0.00     0.00   0  0.00   8  4 88  0.39 0.32 0.33
    0.00   0  0.00     0.00   0  0.00   9  4 87  0.39 0.32 0.33
    0.00   0  0.00     0.00   0  0.00   8  4 88  0.44 0.33 0.33
          disk0           disk1       cpu     load average
    KB/t tps  MB/s     KB/t tps  MB/s  us sy id   1m   5m   15m
    0.00   0  0.00     0.00   0  0.00  11  5 84  0.44 0.33 0.33
    0.00   0  0.00     0.00   0  0.00  13  6 81  0.48 0.34 0.34
    0.00   0  0.00     0.00   0  0.00  10  4 86  0.44 0.34 0.34
    0.00   0  0.00     0.00   0  0.00   8  4 88  0.44 0.34 0.34
    0.00   0  0.00     0.00   0  0.00   8  4 88  0.41 0.33 0.33
    0.00   0  0.00     0.00   0  0.00   8  4 88  0.41 0.33 0.33
    0.00   0  0.00     0.00   0  0.00   8  4 87  0.38 0.32 0.33
    0.00   0  0.00     0.00   0  0.00   7  4 89  0.35 0.32 0.33
    0.00   0  0.00     0.00   0  0.00   8  4 87  0.35 0.32 0.33
    0.00   0  0.00     0.00   0  0.00   7  3 89  0.32 0.31 0.33
    0.00   0  0.00     0.00   0  0.00   8  4 88  0.32 0.31 0.33
    0.00   0  0.00     0.00   0  0.00   7  3 90  0.29 0.31 0.33
  347.11  92 31.29     0.00   0  0.00  20  8 72  0.43 0.34 0.33
   49.98 656 32.03     0.00   0  0.00  32 10 59  0.43 0.34 0.33
  113.45 351 38.90     0.00   0  0.00  40 15 45  0.47 0.35 0.34
  819.41  34 27.20     0.00   0  0.00  26 12 63  0.47 0.35 0.34
  686.99  50 33.76     0.00   0  0.00  18  7 76  0.52 0.36 0.34
  878.17  23 20.01     0.00   0  0.00  20  8 72  0.47 0.35 0.34
    0.00   0  0.00     0.00   0  0.00   8  4 87  0.47 0.35 0.34
    0.00   0  0.00     0.00   0  0.00   8  4 88  0.44 0.34 0.34

You can see in the MB/s column, lots of “0.00″ followed by some bursts of ~30 MB/s. The zeros didn’t actually print at the time, but flooded the screen in bursts when the bottleneck cleared up. It seems to me like it might be some problem with queueing or caching, or maybe the SATA controller on this Mac just isn’t up to the task of SSDs. I’m not sure, but at this point I’m afraid I might have to go back to the 7200 RPM Seagate that came with the Mac. 30-second hangups are far more annoying than having lots of things be slower. Kind of a strange amortization, if you think about it. Anyway, I’ll keep looking into it, now that I know how to reproduce the problem (unzip a huge file).

Updated: A quick Google search for “beach ball mac SSD” turned up this thread which seems to be about this same problem, with a different model SSD. Also referenced in this thread on Apple.com. It feels like the problem may be due to an “old” SSD.

Updated again: Here’s someone having the problem with a Corsair 128 GB SSD: http://forum.corsair.com/v3/showthread.php?t=91061.

Updated again: According to this post, this appears to be a problem with the SATA controller in the 2009 Macbooks. Bah.

]]> http://www.evanhoffman.com/evan/2011/08/23/macbook-pro-locks-up-with-ssd-installed/feed/ 0 I partitioned my laptop so stupidly. http://www.evanhoffman.com/evan/2010/09/20/i-partitioned-my-laptop-so-stupidly/?utm_source=rss&utm_medium=rss&utm_campaign=i-partitioned-my-laptop-so-stupidly http://www.evanhoffman.com/evan/2010/09/20/i-partitioned-my-laptop-so-stupidly/#comments Mon, 20 Sep 2010 15:25:26 +0000 Evan Hoffman http://www.evanhoffman.com/evan/?p=627

When I first installed Linux on my laptop (on my old hard drive) I did it as dual-boot, so I resized my Windows XP partition down to 50 GB, created a 2nd partition for Linux and installed it there. I think I played around with Fedora and Ubuntu and one other distro (maybe FreeBSD?) so I had a bunch of stupid partitions. I eventually went to Linux exclusively and repurposed my XP partition to be my home directory (/docs) and moved all my documents there.

Then I moved to an SSD, which in addition to being incalculably faster than the Seagate Momentus 5400rpm drive, was also bigger – 128 GB instead of 100 GB. This was good, except that my method of moving the data from old drive to new drive was “dd if=/dev/sda of=/dev/sdb”, doing a bit copy from the old drive to the new one. This worked, but it left the old partition table in place on the new drive, basically leaving the extra 28 GB invisible. I didn’t realize this until a couple of weeks ago when I extended the “rear” partition to consume the rest of the space on the disk.

Anyway, now I have this totally stupid partition scheme on the disk:

Retarded partitioning scheme

Retarded partitioning scheme

I could fix it but it seems like a big pain in the ass. Maybe I can clear out /dev/sd8 and copy my docs to it, then copy the root partition to /dev/sda1… Oh, I don’t know. This is why my rule of thumb about partitioning is: don’t.

]]> http://www.evanhoffman.com/evan/2010/09/20/i-partitioned-my-laptop-so-stupidly/feed/ 0 NILFS – A File system to make SSDs scream… in pain? http://www.evanhoffman.com/evan/2009/10/27/nilfs-a-file-system-to-make-ssds-scream-in-pain/?utm_source=rss&utm_medium=rss&utm_campaign=nilfs-a-file-system-to-make-ssds-scream-in-pain http://www.evanhoffman.com/evan/2009/10/27/nilfs-a-file-system-to-make-ssds-scream-in-pain/#comments Tue, 27 Oct 2009 21:05:12 +0000 Evan Hoffman http://www.evanhoffman.com/evan/?p=188

So I got this 128 gig Corsair SSD and put it in my laptop at work. After some fiddling I copied my old disk over to my new disk by booting to Knoppix and doing dd if=/dev/sda of=/dev/sdb bs=4k conv=notrunc,noerror. It’s a lot faster, but what’s really fast now is my Windows XP VM. Anyway, I was looking into other filesystems to try out on SSD to improve speed and I found this article claiming that NILFS is the best choice. So I decided to test it using the same ghetto test I always use for filesystem performance: dd!

Drive info:

Model: ATA CORSAIR CMFSSD-1 (scsi)
Disk /dev/sda: 128GB

The nilfs version in use is whatever’s in yum:

[root@ehoffman ~]# rpm -qai nilfs-utils
Name        : nilfs-utils                  Relocations: (not relocatable)
Version     : 2.0.14                            Vendor: Fedora Project
Release     : 2.fc11                        Build Date: Thu 30 Jul 2009 07:16:08 PM EDT
Install Date: Tue 27 Oct 2009 04:18:28 PM EDT      Build Host: xenbuilder4.fedora.phx.redhat.com
Group       : System Environment/Base       Source RPM: nilfs-utils-2.0.14-2.fc11.src.rpm
Size        : 211949                           License: GPLv2+
Signature   : RSA/8, Thu 30 Jul 2009 09:25:18 PM EDT, Key ID 1dc5c758d22e77f2
Packager    : Fedora Project
URL         : http://www.nilfs.org
Summary     : Utilities for managing NILFS v2 filesystems
Description :
Userspace utilities for creating and mounting NILFS v2 filesystems.

NILFS volume on /dev/sda9, ext3 on /dev/sda8

/dev/sda8 on /docs type ext3 (rw)
/dev/sda9 on /nilfs type nilfs2 (rw,gcpid=3711)

I found this pretty unsettling when I mounted the nilfs volume:

[root@ehoffman ~]# mount -t nilfs2 /dev/sda9 /nilfs/
mount.nilfs2: WARNING! - The NILFS on-disk format may change at any time.
mount.nilfs2: WARNING! - Do not place critical data on a NILFS filesystem.

Write a 100 MB file and a 1.0 GB file to the nilfs volume and the ext3 volume:

[root@ehoffman ~]# dd if=/dev/zero of=/docs/zeros.dat bs=4k count=25600
25600+0 records in
25600+0 records out
104857600 bytes (105 MB) copied, 0.434741 s, 241 MB/s
[root@ehoffman ~]# rm -f /docs/zeros.dat
[root@ehoffman ~]# dd if=/dev/zero of=/docs/zeros.dat bs=4k count=256000
256000+0 records in
256000+0 records out
1048576000 bytes (1.0 GB) copied, 19.6931 s, 53.2 MB/s
[root@ehoffman ~]# rm -f /docs/zeros.dat
[root@ehoffman ~]# dd if=/dev/zero of=/docs/zeros.dat bs=4k count=256000
256000+0 records in
256000+0 records out
1048576000 bytes (1.0 GB) copied, 12.7625 s, 82.2 MB/s
[root@ehoffman ~]# rm -f /docs/zeros.dat
[root@ehoffman ~]# dd if=/dev/zero of=/nilfs/zeros.dat bs=4k count=25600
25600+0 records in
25600+0 records out
104857600 bytes (105 MB) copied, 5.42617 s, 19.3 MB/s
[root@ehoffman ~]# dd if=/dev/zero of=/nilfs/zeros.dat bs=4k count=256000
256000+0 records in
256000+0 records out
1048576000 bytes (1.0 GB) copied, 47.4966 s, 22.1 MB/s
[root@ehoffman ~]# rm -f /nilfs/zeros.dat

With 100 MB, ext3 writes at 200+MB/s and with 1 gig it writes around 50+ MB/s. I think the SATA controller on my laptop is SATA I (not SATA II). On nilfs it seems to hover around 20-25 MB regardless of file size. Anyway, based on this I guess I’ll be staying with ext3/4 for the foreseeable future.

]]> http://www.evanhoffman.com/evan/2009/10/27/nilfs-a-file-system-to-make-ssds-scream-in-pain/feed/ 2