headers
Travis | 2 Aug 2010 05:22
Picon

Re: New to linux- network setup

-----Original Message----- 
From: Meryll Larkin
Sent: Wednesday, July 28, 2010 4:41 PM
To: Johnathan McMenamin ; gslug-general <at> gslug.org
Subject: Re: [Gslug-general] New to linux- network setup

Hi Johnathan,

Even a known distro of Linux (Ubuntu) can be unique and it is difficult to 
know where to start troubleshooting without having a history of the problem.

If you are very new to Linux, and YOU installed your distro of Ubuntu (btw, 
you didn't mention which version), and you didn't have a wired network 
attached to your computer during the installation AND if the installation is 
brand new (you haven't started making lots of configuration changes and 
adding data), you might consider reinstalling from scratch with the network 
cable attached.

A recent version of Ubuntu should find most network cards - and install the 
appropriate drivers - with the exception of those that are newer than the 
version of Ubuntu, and the rare few special-licensing or proprietary network 
cards.

If you installed Ubuntu without attaching a network cable to your computer, 
several networking components which you need might not have been installed 
at all.   With our help, plus the help of a Search Engine and good 
troubleshooting technique, you could probably figure out each one and 
install it from your CD or DVD, or you could reinstall the entire OS and 
save yourself a lot of time.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Fred Morris | 3 Aug 2010 09:03

apropos the Aug. meeting a smallish Perl script

Hi, I signed up on the web site. Here is the referenced script. Then we talk!

--

Fred Morris

--

#!/usr/bin/perl -w

=pod

=head1 dnswatcher.pl -- Collect DNS Answers

This program collects DNS answers (C<CNAME> and C<A> RRs) and writes a file
with statistics about how often the particular name was requested.

You will probably need to run this script as the superuser in order for it
to work.

=head2 Command Modes

B<List available network devices>

  dnswatcher.pl devices

B<Capture traffic for a specified period of time>

  dnswatcher.pl device seconds
(Continue reading)

Permalink | Reply | 7 comments |
headers
Jeremiah Jester | 3 Aug 2010 19:39
Picon

disk io wait help

Hello,

We have a Dell PS4000 SAN and a Poweredge R900 running RHEL 5. Both the SAN and client are running RAID 5 w/ SATA 7200rpm drives. The client is connected to the SAN via iSCSI and performance to the SAN is very quick, verified via dd command. SAN is configured with hardware RAID 5. These are all multi core machines.

When doing a large file copy of about 200GB from the SAN to the client we run into problems. At first the copy works normally but towards the end of the copy waits continously.

xinxiap   8271  8015  0 Aug02 pts/4    00:08:57 cp -fR /vol10/ngsData/runs/100727_HWUSI-EAS1815_00006_FC626R6AAXX .

If i issue an iowait on the SAN mounted disk gives me 100% utilization, cp still in progress.

[root <at> donuts ~]# iostat -x 3 | grep sdd
sdd               1.33     0.67  1.33  0.67   682.67    10.67   346.67     1.00  563.50 499.83  99.97
sdd               3.67     0.67  1.67  0.33   504.00     8.00   256.00     1.22  424.50 499.83  99.97
sdd               1.33     0.00  1.33  0.33   682.67     2.67   411.20     1.08  842.00 600.20 100.03
sdd               3.67     0.33  2.00  0.67   597.33     8.00   227.00     1.00  373.75 375.00 100.00

The local drive/system that we are copying to seems to be pretty much at idle.
 
We only have one user doing a single copy so this doesnt seem to appear to be a limitation of SATA drives as I've done similar copies and more on lower end systems, albeit not over iscsi. Could there be a problem in how this user is running the command via there terminal?
 
Appreciate your help.
 
JJ



--
"Never accept what you hear at face value -- ask questions, do your homework and make sure your interests are protected"

_______________________________________________
Gslug-general mailing list
Gslug-general <at> gslug.org
http://lists.gslug.org/mailman/listinfo/gslug-general
Permalink | Reply | 2 comments |
headers
Ski Kacoroski | 3 Aug 2010 19:42
Picon
Gravatar

Re: disk io wait help


Jeremiah,

I would do:

check the logs on the client for iscsi errors
install sanhq on a client and check for issues with the san.  SanHQ is
Dell's free software for monitoring and debugging san issues.

cheers,

ski

On 08/03/2010 10:39 AM, Jeremiah Jester wrote:
> Hello,
> 
> We have a Dell PS4000 SAN and a Poweredge R900 running RHEL 5. Both the SAN
> and client are running RAID 5 w/ SATA 7200rpm drives. The client is
> connected to the SAN via iSCSI and performance to the SAN is very quick,
> verified via dd command. SAN is configured with hardware RAID 5. These are
> all multi core machines.
> 
> When doing a large file copy of about 200GB from the SAN to the client we
> run into problems. At first the copy works normally but towards the end of
> the copy waits continously.
> 
> xinxiap   8271  8015  0 Aug02 pts/4    00:08:57 cp -fR
> /vol10/ngsData/runs/100727_HWUSI-EAS1815_00006_FC626R6AAXX .
> 
> If i issue an iowait on the SAN mounted disk gives me 100% utilization, cp
> still in progress.
> 
> [root <at> donuts ~]# iostat -x 3 | grep sdd
> sdd               1.33     0.67  1.33  0.67   682.67    10.67   346.67
> 1.00  563.50 499.83  99.97
> sdd               3.67     0.67  1.67  0.33   504.00     8.00   256.00
> 1.22  424.50 499.83  99.97
> sdd               1.33     0.00  1.33  0.33   682.67     2.67   411.20
> 1.08  842.00 600.20 100.03
> sdd               3.67     0.33  2.00  0.67   597.33     8.00   227.00
> 1.00  373.75 375.00 100.00
> 
> The local drive/system that we are copying to seems to be pretty much at
> idle.
> 
> We only have one user doing a single copy so this doesnt seem to appear to
> be a limitation of SATA drives as I've done similar copies and more on lower
> end systems, albeit not over iscsi. Could there be a problem in how this
> user is running the command via there terminal?
> 
> Appreciate your help.
> 
> JJ
> 
> 
> 
> 
> 
> 
> _______________________________________________
> Gslug-general mailing list
> Gslug-general <at> gslug.org
> http://lists.gslug.org/mailman/listinfo/gslug-general

--

-- 
"When we try to pick out anything by itself, we find it
 connected to the entire universe"            John Muir

Chris "Ski" Kacoroski, kacoroski <at> gmail.com, 206-501-9803
or ski98033 on most IM services
Permalink | Reply | 0 comments |
headers
Thushara Wijeratna | 3 Aug 2010 21:04
Picon

Jacob Appelbaum

does anyone know Jacob here - seems he was questioned by the government regarding wikileaks.

_______________________________________________
Gslug-general mailing list
Gslug-general <at> gslug.org
http://lists.gslug.org/mailman/listinfo/gslug-general
Permalink | Reply | 5 comments |
headers
Travis | 4 Aug 2010 00:00
Picon

Re: Jacob Appelbaum

-----Original Message----- 
From: Thushara Wijeratna
Sent: Tuesday, August 03, 2010 12:04 PM
To: gslug-general <at> gslug.org
Subject: [Gslug-general] Jacob Appelbaum

does anyone know Jacob here - seems he was questioned by the government 
regarding wikileaks.

and your point is...........

--

-- 

Travis in Shoreline Washington 

_______________________________________________
Gslug-general mailing list
Gslug-general <at> gslug.org
http://lists.gslug.org/mailman/listinfo/gslug-general
Permalink | Reply | 4 comments |
headers
Thushara Wijeratna | 4 Aug 2010 00:03
Picon

Re: Jacob Appelbaum

just wondered if it is something we, as a community, should know about. and if he is a member of this group, whether we should offer our support in some shape or form.


On Tue, Aug 3, 2010 at 3:00 PM, Travis <travis.eeepc <at> gmail.com> wrote:
-----Original Message----- From: Thushara Wijeratna
Sent: Tuesday, August 03, 2010 12:04 PM
To: gslug-general <at> gslug.org
Subject: [Gslug-general] Jacob Appelbaum


does anyone know Jacob here - seems he was questioned by the government regarding wikileaks.



and your point is...........

--

Travis in Shoreline Washington 

_______________________________________________
Gslug-general mailing list
Gslug-general <at> gslug.org
http://lists.gslug.org/mailman/listinfo/gslug-general
Permalink | Reply | 3 comments |
headers
Lee Colleton | 4 Aug 2010 01:49

Re: Jacob Appelbaum

He wasn't arrested or charged with any crime, why does he need our help? I think the best thing we can do for Jacob would be to support Wikileaks directly, if you're so inclined.

In related news, Bradley Manning has a legal defense fund and his situation is far more dire.

--Lee

On Aug 3, 2010 3:04 PM, "Thushara Wijeratna" <thushw <at> gmail.com> wrote:
> just wondered if it is something we, as a community, should know about. and
> if he is a member of this group, whether we should offer our support in some
> shape or form.
>
>
> On Tue, Aug 3, 2010 at 3:00 PM, Travis <travis.eeepc <at> gmail.com> wrote:
>
>> -----Original Message----- From: Thushara Wijeratna
>> Sent: Tuesday, August 03, 2010 12:04 PM
>> To: gslug-general <at> gslug.org
>> Subject: [Gslug-general] Jacob Appelbaum
>>
>>
>> does anyone know Jacob here - seems he was questioned by the government
>> regarding wikileaks.
>>
>>
>>
>> and your point is...........
>>
>> --
>>
>> Travis in Shoreline Washington
>>

_______________________________________________
Gslug-general mailing list
Gslug-general <at> gslug.org
http://lists.gslug.org/mailman/listinfo/gslug-general
Permalink | Reply | 2 comments |
headers
Chuck Wolber | 4 Aug 2010 02:13
Favicon

Re: Jacob Appelbaum

On Tue, 3 Aug 2010, Lee Colleton wrote:

> In related news, Bradley Manning has a legal defense 
>
fund<http://www.bradleymanning.org/5265/press-release-legal-fund-established-to-fight-imprisonment-of-accused-wikileaks-whistleblower/>and 
> his situation is far more dire.

Yes, and as a corollary, I would like to take the opportunity publically 
get two items off of my chest:

1. I used to have a certain respect for Adrian Lamo.

2. I now believe that he is nothing more than a glory seeking 
script-kiddie loser. I regret any positive feelings I had about him.

..Ch:W..

--

-- 
http://www.quantumlinux.com                 | "An idea does not gain
 Quantum Linux Laboratories, LLC.           |  truth as it gains
 ACCELERATING Business with Open Technology |  followers." Amanda Bloom
_______________________________________________
Gslug-general mailing list
Gslug-general <at> gslug.org
http://lists.gslug.org/mailman/listinfo/gslug-general
Permalink | Reply | 1 comment |
headers
Skylar Thompson | 4 Aug 2010 02:56

Re: disk io wait help


On 08/03/2010 10:39 AM, Jeremiah Jester wrote:
> Hello,
> 
> We have a Dell PS4000 SAN and a Poweredge R900 running RHEL 5. Both the SAN
> and client are running RAID 5 w/ SATA 7200rpm drives. The client is
> connected to the SAN via iSCSI and performance to the SAN is very quick,
> verified via dd command. SAN is configured with hardware RAID 5. These are
> all multi core machines.
> 
> When doing a large file copy of about 200GB from the SAN to the client we
> run into problems. At first the copy works normally but towards the end of
> the copy waits continously.
> 
> xinxiap   8271  8015  0 Aug02 pts/4    00:08:57 cp -fR
> /vol10/ngsData/runs/100727_HWUSI-EAS1815_00006_FC626R6AAXX .
> 
> If i issue an iowait on the SAN mounted disk gives me 100% utilization, cp
> still in progress.
> 
> [root <at> donuts ~]# iostat -x 3 | grep sdd
> sdd               1.33     0.67  1.33  0.67   682.67    10.67   346.67
> 1.00  563.50 499.83  99.97
> sdd               3.67     0.67  1.67  0.33   504.00     8.00   256.00
> 1.22  424.50 499.83  99.97
> sdd               1.33     0.00  1.33  0.33   682.67     2.67   411.20
> 1.08  842.00 600.20 100.03
> sdd               3.67     0.33  2.00  0.67   597.33     8.00   227.00
> 1.00  373.75 375.00 100.00
> 
> The local drive/system that we are copying to seems to be pretty much at
> idle.
> 
> We only have one user doing a single copy so this doesnt seem to appear to
> be a limitation of SATA drives as I've done similar copies and more on lower
> end systems, albeit not over iscsi. Could there be a problem in how this
> user is running the command via there terminal?
> 
> Appreciate your help.

I've run into similar problems, and traced it down to how Linux flushes
out its dirty write buffers. You might try increasing the spread between
the sysctl's vm.dirty_background_ratio and vm.dirty_ratio. The former is
the percentage of physical RAM at which the kernel starts writing out
dirty buffers, while still accepting new writes asynchronously (i.e.
caching). The latter is the point at which the kernel stops accepting
new writes in cache before flushing data to stable storage.

Alternatively, if this is a large-memory system, you might have better
luck just reducing both vm.dirty_background_ratio and vm.dirty_ratio.
This reduces the burstiness of the flushing so it becomes less noticeable.

--

-- 
--
-- Skylar Thompson (skylar <at> cs.earlham.ed)
-- http://www.cs.earlham.edu/~skylar/
Permalink | Reply | 0 comments |

Gmane