Fast, Stable, Cheap – Pick One - Latest Comments

Re: Updating the Firmware on a BCR2000 without the Behringer Update Utility

nexenta — Thu, 30 Jun 2011 04:53:10 -0000

If you are on Mac is more simplistic. Download SysEx Librarian from http://www.snoize.com/SysExLib... (free software) and run on Mac. Download the latest firmware update from Behringer site -> Support -> Download and Drivers section. Just load the downloaded firmware in SysEx Librarian and you will see the file starting to be uploaded on the BCR unit. No need for Midi interface, just with the usb default cable. Also no need for booting in LOAD mode. Hope it helps.

Re: Quickly open the source of a Python module from the command-line

lukecyca — Sat, 08 Jan 2011 11:51:12 -0000

Oops, that was a badly sanitized example, but the problem remains that I can't open a module inside an egg, because the path it returns doesn't exist in the filesystem.


mbp-luke:~ luke$ export EDITOR=`which vim`
mbp-luke:~ luke$ pysource genshi.template.base
Opens a blank vim editor saying:
"/Library/Python/2.6/site-packages/Genshi-0.6-py2.6.egg/genshi/template/base.py" [New DIRECTORY]

or...


mbp-luke:~ luke$ export EDITOR=`which cat`
mbp-luke:~ luke$ pysource genshi.template.base
cat: /Library/Python/2.6/site-packages/Genshi-0.6-py2.6.egg/genshi/template/base.py: Not a directory

Are some editors able to crack open an egg and get at the files inside, as in a path like this?

Re: Quickly open the source of a Python module from the command-line

Kamil Kisiel — Sat, 08 Jan 2011 02:25:39 -0000

Sounds like your $EDITOR isn't set ?

Re: Quickly open the source of a Python module from the command-line

lukecyca — Sat, 08 Jan 2011 00:18:29 -0000

Is there any way to make it work with eggs?

mbp-luke:~ luke$ pysource genshi.template.base
-bash: /Library/Python/2.6/site-packages/Genshi-0.6-py2.6.egg/genshi/template/base.py: Not a directory

Re: Speeding up SQLAlchemy collections with innerjoin

Kamil Kisiel — Fri, 29 Oct 2010 12:56:54 -0000

ORM provides a lot of advantages in providing you domain-specific ways to express your queries, but requires careful tuning of your mapping to get the right performance characteristics. It's in many ways elegant, but as with anything there are trade-offs.

Re: Eclipse, Python, and file extensions

Kamil Kisiel — Fri, 29 Oct 2010 12:42:12 -0000

I've moved back to Vim for my Python development. I discovered how to
integrate pyflakes and made a bunch of other tweaks to my vimrc that
give me nearly the same capabilities as Eclipse but with the superior
editing capabilities of Vim.

Re: Eclipse, Python, and file extensions

Fernando Correia — Fri, 29 Oct 2010 07:02:13 -0000

I wonder how your experiment with PyDev went. I tried most IDEs to work on Python web app projects. In the end I didn't see too much value in them and moved on to Kate. Things were ok for some time but recently I decided to jump into vim, something I thought I would never do, and I'm liking it so far. I don't have much of a save-run cycle, I usually just save a file and alt-tab into Firefox to see the changes.

Re: Speeding up SQLAlchemy collections with innerjoin

Fernando Correia — Fri, 29 Oct 2010 06:57:49 -0000

SQLAlchemy is really well designed and I consider it one of the best ORMs around for any platform. It's great you could tweak it to emit the SQL you wanted. But I wonder if sometimes it wouldn't be cleaner to just write the SELECT statement itself and get done with it... :)

Re: Why Gentoo?

Kamil Kisiel — Sun, 11 Apr 2010 12:35:27 -0000

Well, the learning curve is certainly higher than a Debian or Ubuntu setup. However, the big benefit is that you will learn a lot more about what each component of the system does. If that's your goal, then Gentoo is a good thing to try.

Re: Why Gentoo?

Aerus — Sun, 11 Apr 2010 08:07:07 -0000

Thank you for your post. Every time when I read post about Gentoo, I want to try it. Seriously I used only debian child - Ubuntu Server. Could you please tell if it will be hard for me to use Gentoo ? I've setuped 2 servers on Ubuntu Server distr. Any advice will be appreciated :)

Re: Installing VMWare Tools on Kernel 2.6.22

arlahiru — Wed, 13 Jan 2010 00:36:18 -0000

hi,

I got this error massage when i install vmware-tools 6.5.0

None of the pre-built vmhgfs modules for VMware Tools is suitable for your
running kernel. Do you want this program to try to build the vmhgfs module for
your system (you need to have a C compiler installed on your system)? [yes]

Extracting the sources of the vmhgfs module.

Building the vmhgfs module.

Using 2.6.x kernel build system.
make: Entering directory `/tmp/vmware-config11/vmhgfs-only'
make -C /lib/modules/2.6.31-14-generic/build/include/.. SUBDIRS=$PWD SRCROOT=$PWD/. modules
make[1]: Entering directory `/usr/src/linux-headers-2.6.31-14-generic'
CC [M] /tmp/vmware-config11/vmhgfs-only/backdoor.o
In file included from /tmp/vmware-config11/vmhgfs-only/backdoor.h:29,
from /tmp/vmware-config11/vmhgfs-only/backdoor.c:40:
/tmp/vmware-config11/vmhgfs-only/vm_basic_types.h:108:7: warning: "__FreeBSD__" is not defined
CC [M] /tmp/vmware-config11/vmhgfs-only/backdoorGcc32.o
In file included from /tmp/vmware-config11/vmhgfs-only/backdoor.h:29,
from /tmp/vmware-config11/vmhgfs-only/backdoorGcc32.c:45:
/tmp/vmware-config11/vmhgfs-only/vm_basic_types.h:108:7: warning: "__FreeBSD__" is not defined
CC [M] /tmp/vmware-config11/vmhgfs-only/bdhandler.o
In file included from /tmp/vmware-config11/vmhgfs-only/rpcout.h:30,
from /tmp/vmware-config11/vmhgfs-only/hgfsBd.h:28,
from /tmp/vmware-config11/vmhgfs-only/bdhandler.c:45:
/tmp/vmware-config11/vmhgfs-only/vm_basic_types.h:108:7: warning: "__FreeBSD__" is not defined
In file included from /tmp/vmware-config11/vmhgfs-only/request.h:35,
from /tmp/vmware-config11/vmhgfs-only/bdhandler.c:50:
/tmp/vmware-config11/vmhgfs-only/compat_wait.h:78: error: conflicting types for ‘poll_initwait’
include/linux/poll.h:70: note: previous declaration of ‘poll_initwait’ was here
make[2]: *** [/tmp/vmware-config11/vmhgfs-only/bdhandler.o] Error 1
make[1]: *** [_module_/tmp/vmware-config11/vmhgfs-only] Error 2
make[1]: Leaving directory `/usr/src/linux-headers-2.6.31-14-generic'
make: *** [vmhgfs.ko] Error 2
make: Leaving directory `/tmp/vmware-config11/vmhgfs-only'
Unable to build the vmhgfs module.

The filesystem driver (vmhgfs module) is used only for the shared folder
feature. The rest of the software provided by VMware Tools is designed to work
independently of this feature.

If you wish to have the shared folders feature, you can install the driver by
running vmware-config-tools.pl again after making sure that gcc, binutils, make
and the kernel sources for your running kernel are installed on your machine.
These packages are available on your distribution's installation CD.
[ Press Enter key to continue ]

this is bit similar that error message posted in your article. but not that one

please help me to solve this problem

Re: Installing VMWare Tools on Kernel 2.6.22

CD — Sun, 15 Nov 2009 13:25:04 -0000

Brilliant. It resolve the problem I had with CentOS 5.1

Thanks a lot.

Re: HOWTO: Make a Nagios Dashboard widget in 50 seconds.

mike — Thu, 11 Jun 2009 11:40:17 -0000

Thank you! Didn't know about this Safari feature, and I love nagios.

Re: Updating the Firmware on a BCR2000 without the Behringer Update Utility

Anonymous — Wed, 15 Apr 2009 10:21:00 -0000

This is VERY helpful. Thank you for taking the time to post it.

Love!

Re: Connecting to a Cisco IPSEC VPN from Linux — without the Cisco client

andash — Fri, 06 Mar 2009 19:51:00 -0000

Awesome, thank you for this post. I was just about to compile the Cisco client when I randomly clicked my way to your blog.

Gotta love the internets, thanks again.

Re: Understanding Linux nfsd statistics

Mark Seger — Wed, 11 Feb 2009 19:00:00 -0000

ahh, the old 5 minute rule. my belief is if you're not looking at data very frequently, somewhere between 1 and 10 seconds, it's not worth looking at. This of course assumes you're collectl method doesn't have high overhead. The problem with longer monitoring interval is you end up getting mush! lets say your network is running flat out for a minute and at no load for 4. A 5 minute monitor will say it's 20% loaded and give a very wrong picture.

By default, collectl takes a sample every 10 seconds and only uses about 0.1% of the cpu.

-mark

Re: Understanding Linux nfsd statistics

Kamil Kisiel — Wed, 11 Feb 2009 18:26:04 -0000

@Mark

Looks like a very useful tool. I think it would certainly be useful when high resolution output is desired. We collect SNMP stats about every 5 minutes to avoid overloading the servers with queries. Per-second data would certainly be useful when profiling or benchmarking though.

Re: Understanding Linux nfsd statistics

Mark Seger — Wed, 11 Feb 2009 15:00:00 -0000

re graphing: one of the nifty things about collectl is not only does if report on nfs data, it also reports on just about everything else on your system too AND can provide its output in a format usable by gnuplot or other plotting packages that read space-delimited data.

And if you're really creating, you can even feed the data to ganglia as reported here - www.hp.com/techservers/hpccn/d...

-mark

Re: Understanding Linux nfsd statistics

Kamil Kisiel — Wed, 11 Feb 2009 12:22:10 -0000

@Anonymous above Mark:
No I'm pretty sure it's the number of times all threads were used. The field in the struct in fs/nfsd/nfs3proc.c is "th_fullcnt" and in fs/nfsd/nfssvc.c it's incremented whenever the 10th decile (last value in the histogram) is incremented. In your example, the times for the 10th decile went from 12832.636 to 12832.662, which is easily enough time for the nfsd server to be queried 19 times (the difference in your full-count numbers).

@Mark Seger:
Looks like a nifty tool. I'll have to give it a look some time. Personally I've created a Net-SNMP pass-persist module and have been graphing the values with our network monitoring tool.

Re: Understanding Linux nfsd statistics

Mark Seger — Wed, 11 Feb 2009 07:55:00 -0000

very good article. I've been monitoring a lot of these stats with http://collectl.sourceforge.ne... and can in fact show all the rpc, net and detailed stats as well. Unlink nfsstat which just shows counters, collectl can provide these as rates which I find much more readable/useful.
-mark

Re: Understanding Linux nfsd statistics

Anonymous — Wed, 04 Feb 2009 14:41:00 -0000

useful info. thanks. it seems that the 2nd number in th line is not as you said as number of times ALL threads are in use. see these th lines in two nfsd file 2 seconds apart:
th 8 4646438 725023.990 100118.642 24184.568 0.000 8450.287 4531.316 2778.535 1904.995 0.000 12832.6
36

th 8 4646457 725025.088 100118.680 24184.568 0.000 8450.288 4531.318 2778.536 1904.998 0.000 12832.6
62

it's probably just the number of times ANY thread is used.

my kernel: 2.6.18-8.el5

Re: Understanding Linux nfsd statistics

Valentijn Sessink — Wed, 07 Jan 2009 06:27:00 -0000

Thanks for your excellent article! One small comment: as seen on http://article.gmane.org/gmane... , the "rc" line should have a very low first entry (like 25 in your example), as a cache hit means that a client is retransmitting the same information.

Re: Connecting to a Cisco IPSEC VPN from Linux — without the Cisco client

Anonymous — Fri, 12 Dec 2008 01:14:00 -0000

thanks Kamil. Got it to work on my Ubuntu. I didn't restart after installing so after I disconnected from my VPN, my internet connection went down. But it went away after a restart.

Re: Updating the Firmware on a BCR2000 without the Behringer Update Utility

Andy Chimicles — Wed, 17 Sep 2008 17:26:00 -0000

can't wait to try this tonight. IBehringer's site and documentation sucks for OSX users. Thanks!

Re: AFP + Directory Services on Leopard = Disaster

Luke — Sun, 24 Aug 2008 16:07:00 -0000

We had to downgrade as well back in December, was looking that day to see if anyone had gotten a fix and found out it was solved. Upgraded and haven't had any trouble for a week now. So, I think we're good. Awesome!