Month: December 2012

What many of you probably know is that you can chain commands together with the semicolon. Like so:

unixbench ; echo "unixbench finished" >> unixbench.log

So here we’re running a unixbench, and then directly after, we’re appending “unixbench finished” to unixbench.log.

But what you might not know about is the ‘&&’ replacement for ‘;’. This will only do the next command if the first command returns 0, meaning clean finish.

unixbench && echo "unixbench finished cleanly" >> unixbench.log

And taking it a step further, there’s ‘||’ which does the opposite of ‘&&’. If the first command fails, then the next command happens.

unixbench || echo "unixbench failed" >> unixbench.log

 

As of the Hyper-V Version: 6.2.9200.16384, Ubuntu 12.04 x86_64 performs very slow as a guest os. I switched to CentOS 6.2 x86_64 and performance was dramatically increased. It should be noted that “Linux Integration Services Version 3.4 for Hyper-V” was installed on the CentOS system, but not the Ubuntu system. I’ll investigate this at a later date.

I’ll post a unixbench of both VMs when I get a maintenance window. But for now, examine these load averages. Both running a minecraft server (Bukkit 1.4.6-R0.1) of the same world with the same amount of people. Both had 4GB of ram and 4 CPUs.

Ubuntu 12.04 x86_64: load average: 2.65, 2.92, 2.47

CentOS 6.2 x86_64: load average: 0.14, 0.11, 0.09

—————————–

EDIT 12/27/2012

UnixBench Ubuntu 12.04

   
BYTE UNIX Benchmarks (Version 5.1.3)

   System: u1204vm: GNU/Linux
   OS: GNU/Linux -- 3.2.0-35-generic -- #55-Ubuntu SMP Wed Dec 5 17:42:16 UTC 2012
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6799.9 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 1: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6801.6 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 2: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6801.7 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 3: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6801.4 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   12:26:39 up 3 days, 10:50,  1 user,  load average: 0.31, 0.14, 0.08; runlevel 2

------------------------------------------------------------------------
Benchmark Run: Thu Dec 27 2012 12:26:39 - 12:56:13
4 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       44781167.7 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     3176.1 MWIPS (16.1 s, 7 samples)
Execl Throughput                               2053.4 lps   (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks       1451068.7 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          179747.9 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1079491.1 KBps  (30.0 s, 2 samples)
Pipe Throughput                              889723.3 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 147545.5 lps   (10.0 s, 7 samples)
Process Creation                               4040.6 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   4485.5 lpm   (60.1 s, 2 samples)
Shell Scripts (8 concurrent)                   2245.5 lpm   (60.0 s, 2 samples)
System Call Overhead                        1961736.6 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   44781167.7   3837.3
Double-Precision Whetstone                       55.0       3176.1    577.5
Execl Throughput                                 43.0       2053.4    477.5
File Copy 1024 bufsize 2000 maxblocks          3960.0    1451068.7   3664.3
File Copy 256 bufsize 500 maxblocks            1655.0     179747.9   1086.1
File Copy 4096 bufsize 8000 maxblocks          5800.0    1079491.1   1861.2
Pipe Throughput                               12440.0     889723.3    715.2
Pipe-based Context Switching                   4000.0     147545.5    368.9
Process Creation                                126.0       4040.6    320.7
Shell Scripts (1 concurrent)                     42.4       4485.5   1057.9
Shell Scripts (8 concurrent)                      6.0       2245.5   3742.6
System Call Overhead                          15000.0    1961736.6   1307.8
                                                                   ========
System Benchmarks Index Score                                        1108.3

------------------------------------------------------------------------
Benchmark Run: Thu Dec 27 2012 12:56:13 - 13:25:45
4 CPUs in system; running 4 parallel copies of tests

Dhrystone 2 using register variables      154446771.7 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                    15728.0 MWIPS (13.0 s, 7 samples)
Execl Throughput                               5584.7 lps   (29.6 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks       1143792.2 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          349687.8 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       3281852.5 KBps  (30.0 s, 2 samples)
Pipe Throughput                             8111911.1 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                1301692.0 lps   (10.0 s, 7 samples)
Process Creation                              53932.0 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                  14498.0 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   2048.2 lpm   (60.1 s, 2 samples)
System Call Overhead                       11535804.1 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0  154446771.7  13234.5
Double-Precision Whetstone                       55.0      15728.0   2859.6
Execl Throughput                                 43.0       5584.7   1298.8
File Copy 1024 bufsize 2000 maxblocks          3960.0    1143792.2   2888.4
File Copy 256 bufsize 500 maxblocks            1655.0     349687.8   2112.9
File Copy 4096 bufsize 8000 maxblocks          5800.0    3281852.5   5658.4
Pipe Throughput                               12440.0    8111911.1   6520.8
Pipe-based Context Switching                   4000.0    1301692.0   3254.2
Process Creation                                126.0      53932.0   4280.3
Shell Scripts (1 concurrent)                     42.4      14498.0   3419.3
Shell Scripts (8 concurrent)                      6.0       2048.2   3413.6
System Call Overhead                          15000.0   11535804.1   7690.5
                                                                   ========
System Benchmarks Index Score                                        3936.3

UnixBench CentOS 6.2

   
BYTE UNIX Benchmarks (Version 5.1.3)

   System: x1CenVM: GNU/Linux
   OS: GNU/Linux -- 2.6.32-279.19.1.el6.x86_64 -- #1 SMP Wed Dec 19 07:05:20 UTC 2012
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6799.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSCALL/SYSRET
   CPU 1: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6799.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSCALL/SYSRET
   CPU 2: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6799.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSCALL/SYSRET
   CPU 3: Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz (6799.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSCALL/SYSRET
   02:14:07 up 1 day,  4:57,  2 users,  load average: 0.47, 0.37, 0.21; runlevel 5

------------------------------------------------------------------------
Benchmark Run: Thu Dec 27 2012 02:14:07 - 02:42:36
4 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       43284217.6 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     4547.3 MWIPS (10.3 s, 7 samples)
Execl Throughput                               2993.4 lps   (29.3 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks       1326530.8 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          359629.5 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       3116962.1 KBps  (30.2 s, 2 samples)
Pipe Throughput                             2331204.6 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 471082.1 lps   (10.0 s, 7 samples)
Process Creation                              15342.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   8163.5 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   2658.5 lpm   (60.0 s, 2 samples)
System Call Overhead                        3028768.3 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   43284217.6   3709.0
Double-Precision Whetstone                       55.0       4547.3    826.8
Execl Throughput                                 43.0       2993.4    696.1
File Copy 1024 bufsize 2000 maxblocks          3960.0    1326530.8   3349.8
File Copy 256 bufsize 500 maxblocks            1655.0     359629.5   2173.0
File Copy 4096 bufsize 8000 maxblocks          5800.0    3116962.1   5374.1
Pipe Throughput                               12440.0    2331204.6   1874.0
Pipe-based Context Switching                   4000.0     471082.1   1177.7
Process Creation                                126.0      15342.5   1217.7
Shell Scripts (1 concurrent)                     42.4       8163.5   1925.4
Shell Scripts (8 concurrent)                      6.0       2658.5   4430.8
System Call Overhead                          15000.0    3028768.3   2019.2
                                                                   ========
System Benchmarks Index Score                                        1990.4

------------------------------------------------------------------------
Benchmark Run: Thu Dec 27 2012 02:42:36 - 03:11:13
4 CPUs in system; running 4 parallel copies of tests

Dhrystone 2 using register variables      172485744.7 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                    18251.3 MWIPS (10.1 s, 7 samples)
Execl Throughput                              10103.2 lps   (29.5 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks       1593889.0 KBps  (30.3 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          159649.9 KBps  (30.3 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1614353.2 KBps  (30.0 s, 2 samples)
Pipe Throughput                             2882042.9 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 609999.5 lps   (10.0 s, 7 samples)
Process Creation                              24230.7 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                  11881.5 lpm   (60.2 s, 2 samples)
Shell Scripts (8 concurrent)                   1626.7 lpm   (60.3 s, 2 samples)
System Call Overhead                        3597376.4 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0  172485744.7  14780.3
Double-Precision Whetstone                       55.0      18251.3   3318.4
Execl Throughput                                 43.0      10103.2   2349.6
File Copy 1024 bufsize 2000 maxblocks          3960.0    1593889.0   4025.0
File Copy 256 bufsize 500 maxblocks            1655.0     159649.9    964.7
File Copy 4096 bufsize 8000 maxblocks          5800.0    1614353.2   2783.4
Pipe Throughput                               12440.0    2882042.9   2316.8
Pipe-based Context Switching                   4000.0     609999.5   1525.0
Process Creation                                126.0      24230.7   1923.1
Shell Scripts (1 concurrent)                     42.4      11881.5   2802.3
Shell Scripts (8 concurrent)                      6.0       1626.7   2711.2
System Call Overhead                          15000.0    3597376.4   2398.3
                                                                   ========
System Benchmarks Index Score                                        2706.0

If you have a sudo command you’re trying to run at startup in your /etc/rc.local on a CentOS system, it might not work. The reason is the following error that’s buried in the /var/log/boot.log file:

sudo: sorry, you must have a tty to run sudo

It might be hard to find, but you can always add a ‘sleep 30’ before your sudo command in your /etc/rc.local. This way, the message will show up at the end of the boot.log

Anyhow the reason this is happening is for security. To allow sudo on boot, run ‘visudo’ as root and comment out ‘Defaults requiretty’ as seen below:

# Disable "ssh hostname sudo <cmd>", because it will show the password in clear.
#         You have to run "ssh -t hostname sudo <cmd>".
#
#Defaults    requiretty

Now your sudo command should run just fine at boot time.

——————————

Here’s an example of why you might want to run sudo at boot.

sudo -u minesrv -i /home/minesrv/scripts/startMinecraft

In the above command, I’m running the script startMinecraft, as the user minesrv. startMinecraft has some screen magic, and I want it tied to the minesrv user.

Here’s how you start up a GNU screen session at startup and give it a command to run.

#!/bin/bash
cd /home/minesrv/minecraft_server/
screen -dmS minesrv java -server -Xincgc -Xmx3G -jar minecraft_server.jar nogui

So in this case we’re starting a minecraft server in a screen session.

Per the man page -dm does the following:

-d -m   Start screen in "detached" mode. This creates a new session but
               doesn't  attach  to  it.  This  is  useful  for  system startup
               scripts.

the -S gives the screen session a name so it’s easier to find later on.

Then it runs the command, java, with its parameters.

———————————————————————————

Another useful tool is sudo, so we can run our above startup script as a different user than root.

#!/bin/bash
sudo -u minesrv -i /home/minesrv/scripts/startMinecraft

startMinecraft is the first bash script above. Here we’re using sudo to run it as the user ‘minesrv’.

The -i makes sudo perform all the standard login processes before running the script.

—————————————————————-

And finally, you can just reference the startup script inside /etc/rc.local for most linux distros. That will make the script run at startup.

#!/bin/sh -e
#
# rc.local
#
# This script is executed at the end of each multiuser runlevel.
# Make sure that the script will "exit 0" on success or any other
# value on error.
#
# In order to enable or disable this script just change the execution
# bits.
#
# By default this script does nothing.

/home/minesrv/scripts/startAtBoottimeMinecraft
exit 0