2014.05.21
NANOGのMLでも話題になってたやつ。

アメリカだと選択肢が少ないから eyeball ISP の力が相対的に強いんですかね。


“Chicken” | A Game Played as a Child and by some ISPs with the Internet | Beyond Bandwidth
http://blog.level3.com/global-connectivity/chicken-game-played-child-isps-internet/
要約「Comcastがpeerじゃなくて金払えとか言って来るのマジむかつく。お客さーん、ビデオが止まるのはお宅のプロバイダーのせいですよー。」(注: Level3 はCDN事業者でもある)


Observations of an Internet Middleman | Beyond Bandwidth
http://blog.level3.com/global-connectivity/observations-internet-middleman/
要約「俺らまじデカイし(peerとか合計13Tbpsだし)、すげー金かけてるからお客さんがインターネット使えてるし、だからComcastムカつく」


Level3 is without peer, now what to do?
http://www.cringely.com/2014/05/06/14890/
要約「最近はややこしいから古き良き時代に戻れるといいののね」

An 'internet middleman' calls out six ISPs for letting connections slow down
http://www.engadget.com/2014/05/05/level-3-internet-middlemen/
要約「ガジェット好き」

結は知らない。
2013.07.11
しばらく前から FreeBSD 9.1-RELEASE-p4 の箱がリブートしない(All buffers synced.
の後で止まっちゃう)ので困ってた。(-p4 より以前から。たぶん、9.1 になった時から)
root@host:/root # reboot
Jul 10 10:06:27 host reboot: rebooted by root
Jul 10 10:06:27 host syslogd: exiting on signal 15
Waiting (max 60 seconds) for system process `vnlru' to stop...done
Waiting (max 60 seconds) for system process `bufdaemon' to stop...done
Waiting (max 60 seconds) for system process `syncer' to stop...
Syncing disks, vnodes remaining...0 0 0 done
All buffers synced.
Uptime: 8h1m32s
usbus0: Controller shutdown
uhub0: at usbus0, port 1, addr 1 (disconnected)
usbus0: Controller shutdown complete
usbus1: Controller shutdown
uhub1: at usbus1, port 1, addr 1 (disconnected)
usbus1: Controller shutdown complete
usbus2: Controller shutdown
こんな感じで止まってる。

てっきりハードウェアの問題かと思って放っておいたんだけど、ふと思いついてちょっと "freebsd reboot stall usbus
controller shutdown" とかでググってみたところ、kern/172952: 9.1
hangs on reboot after all buffers synced
が見つかった。これによると 9.1 で ZFS
な箱が同じようにリブートしないという問題らしい。ログを読んでいくとパッチが提案されており、9-stable にすでに commit されてるらしい。

そこで、さくっと svnup stable して、9-stable に上げてみた。(途中 xinstall
のコンパイルに失敗するとか関係ない穴にはまったけど、src 全体を svnup で取り直したらなおった。)

期待を込めて(複数回)リブートしたけど、まだダメだった。
root@host:~ # reboot
Jul 11 22:26:59 host reboot: rebooted by root
Jul 11 22:27:00 host syslogd: exiting on signal 15
Waiting (max 60 seconds) for system process `vnlru' to stop...done
Waiting (max 60 seconds) for system process `bufdaemon' to stop...done
Waiting (max 60 seconds) for system process `syncer' to stop...
Syncing disks, vnodes remaining...0 0 0 done
All buffers synced.
Uptime: 9m8s
usbus0: Controller shutdown
uhub0: at usbus0, port 1, addr 1 (disconnected)
usbus0: Controller shutdown complete
usbus1: Controller shutdown
uhub1: at usbus1, port 1, addr 1 (disconnected)
usbus1: Controller shutdown complete
usbus2: Controller shutdown
uhub2: at usbus2, port 1, addr 1 (disconnected)
ugen2.2: at usbus2 (disconnected)
umass0: at uhub2, port 3, addr 2 (disconnected)
(da0:umass-sim0:0:0:0): lost device - 0 outstanding, 3
refs
こんな感じで止まっちゃう。メッセージは変わったけど、相変わらずリブートしない。

そこで、再度 "freebsd zfs usb reboot failure" とかでググったところ、Problem with shutdown and
external storage
のスレッドが見つかり、最後のポストにある
hw.usb.no_shutdown_wait=1
を試したところ、ビンゴ! ついに無事リブートするようになった。

9.1 にもなってソフトウェアの問題とは思いもしなかった(しかも、確か 9.0 の時は問題は起こってなかった)ので、ここに到達するまで時間がかかった。

すっきり。
2013.07.10
USB外付けディスク ( WD
Elements 2TB Desktop External Hard Drive
) をZFSで使ってると、しょっちゅう CHECKSUM
エラーが報告されるから、何も根拠はないけど、以下を追加してカーネルを入れ替えてみた。

/usr/src/sys/dev/usb/usbdevs:
product WESTERN ELEMENTS 0x1021 Elements External HDD

/usr/src/sys/dev/usb/quirk/usb_quirk.c:
USB_QUIRK(WESTERN, ELEMENTS, 0x0000, 0xffff, UQ_MSC_NO_SYNC_CACHE),
2012.12.05
done.
2012.12.03
--- com.apple.netbiosd.plist.orig       2012-12-03 11:02:23.000000000 +0900
+++ com.apple.netbiosd.plist 2012-11-27 15:52:56.000000000 +0900
@@ -3,9 +3,9 @@
<plist version="1.0">
<dict>
<key>EnableTransactions</key>
- <true/>
- <key>Disabled</key>
<false/>
+ <key>Disabled</key>
+ <true/>
<key>Label</key>
<string>com.apple.netbiosd</string>
<key>ProgramArguments</key>
And
# sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.netbiosd.plist
2012.11.05
■ 毎月の請求
1,960基本使用料 (プランZシンプル)
-980誰でも割
5,460ISフラット
315IS NET
0分割支払金
-2,140毎月割
-980男子割 (MNP)
3,635合計


■ 請求書上での表記
914(1960 - 980)/1.05-20 (基本使用料 - 誰でも割 - Web請求)
5,500(5460+315)/1.05
-2,972(毎月割 + 男子割)/1.05

2012.11.05
■ 毎月の請求
1,960基本使用料 (LTEプラン)
-980誰でも割
5,460LTEフラット
315LTE NET
2,570分割支払金
-2,570毎月割
-980男子割 (MNP)
5,775合計


■ 請求書上での表記
934(1960 - 980)/1.05 (基本使用料 - 誰でも割)
5,500(5460+315)/1.05
-3,381(毎月割 + 男子割)/1.05
2012.08.25
Connecting Cisco 2600/2800/3600 async (CAB-OCTAL-ASYNC) to BayTech RPC3 console port.

It needs to use specially crafted cable. Pinouts on each end is explained here.
http://www.baytech.net/support/cable_adapters.shtml
http://www.cisco.com/en/US/tech/tk801/tk36/technologies_tech_note09186a00801c9a10.shtml

Basically what you need is this.
http://www.baytech.net/cgi-bin/showprod?prod=RJ08X007-07

If you're using standard Ethernet cable which looks like this:

(left to right)
1 Orange white
2 Orange
3 Green white
4 Blue
5 Blue white
6 Green
7 Brown white
8 Brown

Then the other end (which connects to RPC3) should look like this:

(left to right)
1 Orange
2 Blue white
3 Orange white
4 Green white
5 Green
6 Brown
7 Blue
8 Brown white

Which means:

1 - 7 (Brown white)
2 - 4 (Blue)
3 - 8 (Brown)
4 - 6 (Green)
5 - 3 (Green white)
6 - 1 (Orange white)
7 - 5 (Blue white)
8 - 2 (Orange)

cisco-to-rpc.png
2012.07.31
Set "PermitRootLogin forced-commands-only" in /etc/ssh/sshd_config and,
command="rsync --server -vulogDtpre.iLsf --delete . /backup",no-pty,no-agent-forwarding,no-port-forwarding ssh-rsa ssh-rsa xxxxx
in ~root/.ssh/authorized_keys.
2012.07.28
http://rewse.jp/blog/p/3283 の seqread.fio と同じテストをやってみた。

設定ファイルはこう。
[seqread]
readwrite=read
blocksize=1m
size=1g
directory=/root
direct=1
loops=5
結果はこんな感じ。

TargetbwIOPS
内蔵ディスク1129 MB/s125
内蔵ディスク296 MB/s93
外付け(USB2.0)36 MB/s35
お名前.com (virtio)326 MB/s318
お名前.com203 MB/s198
ServersMan@VPS2 MB/s1

元のページの方の結果はこの通り。
seqread: (groupid=0, jobs=1): err= 0: pid=11993
read : io=5120.0MB, bw=128587KB/s, iops=125 , runt= 40773msec
clat (msec): min=6 , max=201 , avg= 7.95, stdev= 5.06
lat (msec): min=6 , max=201 , avg= 7.95, stdev= 5.06
bw (KB/s) : min=46916, max=148317, per=100.45%, avg=129168.97, stdev=18220.10
cpu : usr=0.00%, sys=0.13%, ctx=5173, majf=0, minf=283
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued r/w/d: total=5120/0/0, short=0/0/0

lat (msec): 10=95.55%, 20=1.04%, 50=3.38%, 250=0.04%

Run status group 0 (all jobs):
READ: io=5120.0MB, aggrb=128587KB/s, minb=131673KB/s, maxb=131673KB/s, mint=40773msec, maxt=40773msec

Disk stats (read/write):
sdd: ios=11566/204, merge=12732/7, ticks=61892/38088, in_queue=95940, util=97.95%

同じディスクを使って FreeBSD 9.0 上で行なった結果はこの通り。当たり前だけどほとんど変わらない。
ada0 at ata2 bus 0 scbus1 target 0 lun 0
ada0: <ST2000DL003-9VT166 CC3C> ATA-8 SATA 3.x device
ada0: 150.000MB/s transfers (SATA 1.x, UDMA5, PIO 8192bytes)
ada0: 1907729MB (3907029168 512 byte sectors: 16H 63S/T 16383C)
ada0: Previously was known as ad4
seqread: (g=0): rw=read, bs=1M-1M/1M-1M, ioengine=sync, iodepth=1
2.0.8
Starting 1 process
seqread: Laying out IO file(s) (1 file(s) / 1024MB)
Jobs: 1 (f=1): [R] [100.0% done] [104.4M/0K /s] [104 /0 iops] [eta 00m:00s]
seqread: (groupid=0, jobs=1): err= 0: pid=100860
read : io=5120.0MB, bw=123650KB/s, iops=120 , runt= 42401msec
clat (usec): min=479 , max=61769 , avg=8275.20, stdev=4561.82
lat (usec): min=479 , max=61770 , avg=8275.67, stdev=4561.88
clat percentiles (usec):
| 1.00th=[ 502], 5.00th=[ 516], 10.00th=[ 580], 20.00th=[ 8096],
| 30.00th=[ 8640], 40.00th=[ 9536], 50.00th=[ 9664], 60.00th=[10176],
| 70.00th=[10176], 80.00th=[10304], 90.00th=[10688], 95.00th=[10688],
| 99.00th=[19840], 99.50th=[20864], 99.90th=[44800], 99.95th=[51456],
| 99.99th=[61696]
bw (KB/s) : min=90934, max=1552384, per=95.78%, avg=118435.81, stdev=161443.26
lat (usec) : 500=0.74%, 750=18.93%, 1000=0.31%
lat (msec) : 10=37.38%, 20=41.80%, 50=0.76%, 100=0.08%
cpu : usr=0.14%, sys=13.09%, ctx=37013, majf=0, minf=261
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=5120/w=0/d=0, short=r=0/w=0/d=0

Run status group 0 (all jobs):
READ: io=5120.0MB, aggrb=123649KB/s, minb=123649KB/s, maxb=123649KB/s, mint=42401msec, maxt=42401msec

もう一個のハードディスク。少し遅い。
ada1 at ata3 bus 0 scbus2 target 1 lun 0
ada1: <SAMSUNG HD103UJ 1AA01118> ATA-7 SATA 2.x device
ada1: 150.000MB/s transfers (SATA 1.x, UDMA5, PIO 8192bytes)
ada1: 953869MB (1953525168 512 byte sectors: 16H 63S/T 16383C)
ada1: Previously was known as ad7
seqread: Laying out IO file(s) (1 file(s) / 1024MB)
Jobs: 1 (f=1): [R] [100.0% done] [77436K/0K /s] [75 /0 iops] [eta 00m:00s]
seqread: (groupid=0, jobs=1): err= 0: pid=103259
read : io=5120.0MB, bw=96067KB/s, iops=93 , runt= 54575msec
clat (usec): min=472 , max=58292 , avg=10652.27, stdev=5483.74
lat (usec): min=472 , max=58293 , avg=10652.84, stdev=5483.83
clat percentiles (usec):
| 1.00th=[ 490], 5.00th=[ 516], 10.00th=[ 556], 20.00th=[10048],
| 30.00th=[12224], 40.00th=[12352], 50.00th=[12352], 60.00th=[12352],
| 70.00th=[12480], 80.00th=[13888], 90.00th=[14016], 95.00th=[14016],
| 99.00th=[22400], 99.50th=[22400], 99.90th=[38144], 99.95th=[38144],
| 99.99th=[58112]
bw (KB/s) : min=73142, max=1605632, per=96.22%, avg=92434.55, stdev=149105.32
lat (usec) : 500=1.41%, 750=18.52%, 1000=0.06%
lat (msec) : 20=77.05%, 50=2.95%, 100=0.02%
cpu : usr=0.09%, sys=10.21%, ctx=38148, majf=0, minf=261
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=5120/w=0/d=0, short=r=0/w=0/d=0

Run status group 0 (all jobs):
READ: io=5120.0MB, aggrb=96067KB/s, minb=96067KB/s, maxb=96067KB/s, mint=54575msec, maxt=54575msec

お名前.com VPS(KVM)(2GBプラン)上で virtio を有効にした上で行なったのがこれ。3倍弱速い。すごく速いです。
vtblk0: <VirtIO Block Adapter> on virtio_pci1
vtblk0: 204800MB (419430400 512 byte sectors)
seqread: (g=0): rw=read, bs=1M-1M/1M-1M, ioengine=sync, iodepth=1
2.0.8
Starting 1 process
seqread: Laying out IO file(s) (1 file(s) / 1024MB)
Jobs: 1 (f=1): [R] [100.0% done] [479.7M/0K /s] [479 /0 iops] [eta 00m:00s]
seqread: (groupid=0, jobs=1): err= 0: pid=100517
read : io=5120.0MB, bw=325807KB/s, iops=318 , runt= 16092msec
clat (usec): min=260 , max=73811 , avg=3092.02, stdev=3514.33
lat (usec): min=269 , max=73820 , avg=3103.96, stdev=3514.25
clat percentiles (usec):
| 1.00th=[ 318], 5.00th=[ 1624], 10.00th=[ 1688], 20.00th=[ 1752],
| 30.00th=[ 1816], 40.00th=[ 1896], 50.00th=[ 1976], 60.00th=[ 2064],
| 70.00th=[ 2224], 80.00th=[ 2736], 90.00th=[ 6112], 95.00th=[10560],
| 99.00th=[18816], 99.50th=[21632], 99.90th=[31616], 99.95th=[36608],
| 99.99th=[74240]
bw (KB/s) : min=79712, max=557056, per=92.45%, avg=301222.97, stdev=149712.28
lat (usec) : 500=1.50%
lat (msec) : 2=52.07%, 4=33.42%, 10=7.23%, 20=5.00%, 50=0.76%
lat (msec) : 100=0.02%
cpu : usr=0.34%, sys=15.72%, ctx=7032, majf=4, minf=262
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=5120/w=0/d=0, short=r=0/w=0/d=0

Run status group 0 (all jobs):
READ: io=5120.0MB, aggrb=325806KB/s, minb=325806KB/s, maxb=325806KB/s, mint=16092msec, maxt=16092msec
virtio使わなかったらこう。virtio使った時より30%くらい遅い。
ada0 at ata1 bus 0 scbus1 target 0 lun 0
ada0: <QEMU HARDDISK 0.15.0> ATA-7 device
ada0: 16.700MB/s transfers (WDMA2, PIO 8192bytes)
ada0: 204800MB (419430400 512 byte sectors: 16H 63S/T 16383C)
ada0: Previously was known as ad2
seqread: (groupid=0, jobs=1): err= 0: pid=100401
read : io=5120.0MB, bw=202866KB/s, iops=198 , runt= 25844msec
clat (msec): min=2 , max=85 , avg= 5.01, stdev= 5.80
lat (msec): min=2 , max=85 , avg= 5.02, stdev= 5.80
clat percentiles (usec):
| 1.00th=[ 2480], 5.00th=[ 2576], 10.00th=[ 2672], 20.00th=[ 2864],
| 30.00th=[ 3088], 40.00th=[ 3568], 50.00th=[ 3792], 60.00th=[ 3888],
| 70.00th=[ 3984], 80.00th=[ 4192], 90.00th=[ 7392], 95.00th=[13760],
| 99.00th=[33024], 99.50th=[43264], 99.90th=[61696], 99.95th=[77312],
| 99.99th=[85504]
bw (KB/s) : min=80000, max=299008, per=96.87%, avg=196517.25, stdev=51283.94
lat (msec) : 4=70.98%, 10=21.66%, 20=4.57%, 50=2.50%, 100=0.29%
cpu : usr=0.17%, sys=7.35%, ctx=7767, majf=4, minf=276
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=5120/w=0/d=0, short=r=0/w=0/d=0

Run status group 0 (all jobs):
READ: io=5120.0MB, aggrb=202866KB/s, minb=202866KB/s, maxb=202866KB/s, mint=25844msec, maxt=25844msec

ついでに ServersMan@VPS(Entry)でも実行した。すごく遅いです。
seqread: (groupid=0, jobs=1): err= 0: pid=627
read : io=1024.0MB, bw=2023.4KB/s, iops=1 , runt=2591595msec
clat (msec): min=9 , max=5747 , avg=506.10, stdev=609.76
lat (msec): min=9 , max=5747 , avg=506.10, stdev=609.76
clat percentiles (msec):
| 1.00th=[ 41], 5.00th=[ 99], 10.00th=[ 111], 20.00th=[ 153],
| 30.00th=[ 184], 40.00th=[ 210], 50.00th=[ 212], 60.00th=[ 241],
| 70.00th=[ 302], 80.00th=[ 996], 90.00th=[ 1696], 95.00th=[ 1696],
| 99.00th=[ 1795], 99.50th=[ 1795], 99.90th=[ 2180], 99.95th=[ 2245],
| 99.99th=[ 5735]
bw (KB/s) : min= 178, max=11861, per=100.00%, avg=3141.50, stdev=2616.53
lat (msec) : 10=0.08%, 20=0.23%, 50=0.96%, 100=5.04%, 250=55.68%
lat (msec) : 500=17.40%, 750=0.43%, 1000=0.18%, 2000=19.84%, >=2000=0.16%
cpu : usr=0.01%, sys=0.10%, ctx=44149, majf=0, minf=0
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=5120/w=0/d=0, short=r=0/w=0/d=0

Run status group 0 (all jobs):
READ: io=5120.0MB, aggrb=2023KB/s, minb=2023KB/s, maxb=2023KB/s, mint=2591595msec, maxt=2591595msec

外付けのUSB2.0ハードディスクだとこんな感じ。
da0 at umass-sim0 bus 0 scbus4 target 0 lun 0
da0: <WD Ext HDD 1021 2021> Fixed Direct Access SCSI-4 device
da0: 40.000MB/s transfers
da0: 1907727MB (3907024896 512 byte sectors: 255H 63S/T 243201C)
seqread: (groupid=0, jobs=1): err= 0: pid=103304
read : io=5120.0MB, bw=36294KB/s, iops=35 , runt=144455msec
clat (usec): min=509 , max=642934 , avg=28202.87, stdev=25625.75
lat (usec): min=510 , max=642934 , avg=28203.62, stdev=25625.79
clat percentiles (usec):
| 1.00th=[ 620], 5.00th=[ 740], 10.00th=[ 852], 20.00th=[15040],
| 30.00th=[29568], 40.00th=[29824], 50.00th=[29824], 60.00th=[29824],
| 70.00th=[30080], 80.00th=[30080], 90.00th=[41216], 95.00th=[53504],
| 99.00th=[95744], 99.50th=[185344], 99.90th=[358400], 99.95th=[411648],
| 99.99th=[643072]
bw (KB/s) : min= 3043, max=954368, per=90.51%, avg=32850.75, stdev=56253.15
lat (usec) : 750=5.31%, 1000=8.89%
lat (msec) : 2=5.70%, 4=0.08%, 20=0.04%, 50=74.41%, 100=4.59%
lat (msec) : 250=0.76%, 500=0.20%, 750=0.02%
cpu : usr=0.09%, sys=5.49%, ctx=56527, majf=0, minf=261
IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=5120/w=0/d=0, short=r=0/w=0/d=0

Run status group 0 (all jobs):
READ: io=5120.0MB, aggrb=36294KB/s, minb=36294KB/s, maxb=36294KB/s, mint=144455msec, maxt=144455msec
Rebuild at 2016/12/08 03:44
×

この広告は1年以上新しい記事の投稿がないブログに表示されております。