FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Debian > Debian User

 
 
LinkBack Thread Tools
 
Old 05-12-2011, 02:17 AM
Aldyth Maharsha
 
Default Help About Squeeze

Hi list i'm sorry if my english too bad :-)

I'm using debian squeeze 2.6.32-5-686 #1 SMP running in machine AMD Sempron(tm)** 2800+ with 1GB DDR
I'm having trouble, my debian squeeze i'm using for samba file sharing and 200 user access that. My user using server for data sharing, myob, etc.

My trouble is my server like "sleep" and if it's sleep i cannot ping, ssh and anything from another computer, to "wakeup" my server i must hit any button in the keyboard(server keyboard) and i can ping again and after 10-15 minutes it is sleep again. I'm using two harddisk, 80GB(system and home partition but system and home partition located at different partition), 160GB(user data sharing like office file, accounting file).


At second harddisk(160GB) have partition error but i can't using fsck because if i'm running fsck force it is can delete the important data because my backup server not running(down) and i must backup manually. My log file like syslog and kern.log did not show any error, it is running like my system is normal, i'm check with lsof +D /var/log my ryslogd runs well. My question is what causes my debian squeeze act like it?, i'm using debian from sarge, etch, lenny and i have never encountered this problem and i'm believe debian still best linux distribution i'm ever have..:-)


Any idea list?, thanks before for helping

Best Regards,
Aldyth M
 
Old 05-12-2011, 07:14 AM
Markus Neviadomski
 
Default Help About Squeeze

Hi,

Thats sound like a fake, sorry. A single-core CPU with a some piece of
RAM as file server for 200 users? No...

Buy a new hardware with 4 or 6 cores, 16GB RAM and a raid controller and
some disks. Then copy your data from the old system on the new raid
system and everything is fine.


Am 12.05.2011 04:17, schrieb Aldyth Maharsha:
> Hi list i'm sorry if my english too bad :-)
>
> I'm using debian squeeze 2.6.32-5-686 #1 SMP running in machine AMD
> Sempron(tm) 2800+ with 1GB DDR
> I'm having trouble, my debian squeeze i'm using for samba file sharing
> and 200 user access that. My user using server for data sharing, myob,
> etc.
> My trouble is my server like "sleep" and if it's sleep i cannot ping,
> ssh and anything from another computer, to "wakeup" my server i must
> hit any button in the keyboard(server keyboard) and i can ping again
> and after 10-15 minutes it is sleep again. I'm using two harddisk,
> 80GB(system and home partition but system and home partition located
> at different partition), 160GB(user data sharing like office file,
> accounting file).
>
> At second harddisk(160GB) have partition error but i can't using fsck
> because if i'm running fsck force it is can delete the important data
> because my backup server not running(down) and i must backup manually.
> My log file like syslog and kern.log did not show any error, it is
> running like my system is normal, i'm check with lsof +D /var/log my
> ryslogd runs well. My question is what causes my debian squeeze act
> like it?, i'm using debian from sarge, etch, lenny and i have never
> encountered this problem and i'm believe debian still best linux
> distribution i'm ever have..:-)
>
> Any idea list?, thanks before for helping
>
> Best Regards,
> Aldyth M


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4DCB88E8.4010309@dieitexperten.de">http://lists.debian.org/4DCB88E8.4010309@dieitexperten.de
 
Old 05-12-2011, 07:40 AM
Thierry Chatelet
 
Default Help About Squeeze

On Thursday 12 May 2011 09:14:48 Markus Neviadomski wrote:
> Hi,
>
> Thats sound like a fake, sorry. A single-core CPU with a some piece of
> RAM as file server for 200 users? No...
>
> Buy a new hardware with 4 or 6 cores, 16GB RAM and a raid controller and
> some disks. Then copy your data from the old system on the new raid
> system and everything is fine.
>

Big help you are giving here! Maybe money is not the most common thing in
Aldyth's country, and resources are used to their last extend.

So, Aldyth, are you running any desktop like kde, gnome? On my pc kde power
managment set up itself by default to put everything to sleep after some time.
I had to reset everything to previous setting. By the way, I think there is way
for reporting a bug against powermanagment because it reset the settings every
time it is updated.
Thierry


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 201105120940.20407.tchatelet@free.fr">http://lists.debian.org/201105120940.20407.tchatelet@free.fr
 
Old 05-12-2011, 07:55 AM
Jeroen van Aart
 
Default Help About Squeeze

Markus Neviadomski wrote:

Thats sound like a fake, sorry. A single-core CPU with a some piece of
RAM as file server for 200 users? No...


I think a 2800 Sempron will have no trouble serving files, even to 200
people. Unless they all simultaneously started copying 10+ GB of data
around on the server itself. But in my experience typical file server
activity is sporadic with occasional bursts.


Besides the bottleneck with a file server normally is not CPU
performance but network performance. You'd likely saturate the network
bandwidth before you'd ever hog the CPU.



Am 12.05.2011 04:17, schrieb Aldyth Maharsha:

My trouble is my server like "sleep" and if it's sleep i cannot ping,
ssh and anything from another computer, to "wakeup" my server i must


Sounds like you have some power manager that's configured to put the
system to sleep. I assume you have some desktop environment running on
it. Try to access its power management configuration editor and turn off
anything that makes it sleep after a certain time.


In gnome it's called "power management" in the
debian->system->preferences menu.


Greetings,
Jeroen

--
http://goldmark.org/jeff/stupid-disclaimers/
http://linuxmafia.com/~rick/faq/plural-of-virus.html


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Archive: 4DCB9268.3040300@mompl.net">http://lists.debian.org/4DCB9268.3040300@mompl.net
 
Old 05-12-2011, 08:29 AM
Markus Neviadomski
 
Default Help About Squeeze

Am 12.05.2011 09:40, schrieb Thierry Chatelet:
> On Thursday 12 May 2011 09:14:48 Markus Neviadomski wrote:
>> Hi,
>>
>> Thats sound like a fake, sorry. A single-core CPU with a some piece of
>> RAM as file server for 200 users? No...
>>
>> Buy a new hardware with 4 or 6 cores, 16GB RAM and a raid controller and
>> some disks. Then copy your data from the old system on the new raid
>> system and everything is fine.
>>
> Big help you are giving here! Maybe money is not the most common thing in
> Aldyth's country, and resources are used to their last extend.

He wrote something about a partly broken harddisk or file system.
However, he has to buy some new disks. And by that, he could establish a
small software raid or use the big solution, a hardware raid controller.
Fileserver w/o backup and w/o working raid cause sleepless nights.

And the costs of hardware could be igored, if you have to recover broken
or lost data w/o raid or backup!

regards,
Markus
> So, Aldyth, are you running any desktop like kde, gnome? On my pc kde power
> managment set up itself by default to put everything to sleep after some time.
> I had to reset everything to previous setting. By the way, I think there is way
> for reporting a bug against powermanagment because it reset the settings every
> time it is updated.
> Thierry
>
>


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4DCB9A62.6060809@dieitexperten.de">http://lists.debian.org/4DCB9A62.6060809@dieitexperten.de
 
Old 05-12-2011, 08:58 AM
Stan Hoeppner
 
Default Help About Squeeze

On 5/12/2011 2:14 AM, Markus Neviadomski wrote:

Hi,

Thats sound like a fake, sorry. A single-core CPU with a some piece of
RAM as file server for 200 users? No...

Buy a new hardware with 4 or 6 cores, 16GB RAM and a raid controller and
some disks. Then copy your data from the old system on the new raid
system and everything is fine.


You didn't answer the OP's question. And your upgrade recommendation is
complete overkill. A modern ~2Ghz Sempron can easily saturate a GbE
pipe without using jumbo frames. 1GB of RAM is plenty for serving 200
office environment Samba clients--512MB would even be sufficient. If
he's lacking performance it's due to insufficient head seeks bandwidth.
He didn't mention a capacity shortage and he currently has ~250GB of
disk. A single 250GB SSD would solve that problem instantly for
$400-600 USD with an MLC drive. Two mirrored would be better, so
$400-1200 USD. This would yield 50,000 seeks/second vs the 300 he has
now, and ~250MB/s bandwidth, vs the ~160MB/s he currently has.




Am 12.05.2011 04:17, schrieb Aldyth Maharsha:

Hi list i'm sorry if my english too bad :-)

I'm using debian squeeze 2.6.32-5-686 #1 SMP running in machine AMD
Sempron(tm) 2800+ with 1GB DDR
I'm having trouble, my debian squeeze i'm using for samba file sharing
and 200 user access that. My user using server for data sharing, myob,
etc.
My trouble is my server like "sleep" and if it's sleep i cannot ping,
ssh and anything from another computer, to "wakeup" my server i must
hit any button in the keyboard(server keyboard) and i can ping again
and after 10-15 minutes it is sleep again. I'm using two harddisk,
80GB(system and home partition but system and home partition located
at different partition), 160GB(user data sharing like office file,
accounting file).

At second harddisk(160GB) have partition error but i can't using fsck
because if i'm running fsck force it is can delete the important data
because my backup server not running(down) and i must backup manually.
My log file like syslog and kern.log did not show any error, it is
running like my system is normal, i'm check with lsof +D /var/log my
ryslogd runs well. My question is what causes my debian squeeze act
like it?, i'm using debian from sarge, etch, lenny and i have never
encountered this problem and i'm believe debian still best linux
distribution i'm ever have..:-)

Any idea list?, thanks before for helping


Yes. Turn off all power saving features in the system BIOS. A headless
installation has no power saving by default AFAIK, so apparently you're
running a GUI desktop. Find the power management application in one of
the control panels and disable all power saving features. They only
cause headaches on servers, as you've discovered. Unless you truly
*need* a GUI on your server console, get rid of it. If you need a GUI
to manage a Linux server then I'd say you really need to sharpen your
admin skill set.


Best of luck.

--
Stan


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Archive: 4DCBA149.1080605@hardwarefreak.com">http://lists.debian.org/4DCBA149.1080605@hardwarefreak.com
 
Old 05-12-2011, 09:20 AM
Stan Hoeppner
 
Default Help About Squeeze

On 5/12/2011 3:29 AM, Markus Neviadomski wrote:


And the costs of hardware could be igored, if you have to recover broken
or lost data w/o raid or backup!


This typically holds true in the US where the total cost of labor is far
greater than hardware. This is definitely not the case in 'developing'
countries, where the acquisition cost of a single tier 1 server may very
well be greater than an SA's yearly salary.


~$20k can buy you a 2 socket 24 core AMD Magny Cours HP server with 32GB
RAM, quad GbE ports, a 10 GbE PCIe x4/x8 NIC, LSI's top of the line PCIe
x8 RAID HBA with 1GB BBWC and 2 SFF8088 SAS ports, two LSI 24 drive 2.5"
chainable SAS enclosures w/ internal expanders and 48 SAS drives of
300GB capacity and 15k spindle speed.


The cost of this system, in many parts of the world, may be double (or
more) the yearly salary of the SA managing it. In the US this system's
price tag will equal about 1/4 to 1/6th the SA's total yearly
compensation package, depending on the SA's state/city.


--
Stan


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Archive: 4DCBA66C.6020905@hardwarefreak.com">http://lists.debian.org/4DCBA66C.6020905@hardwarefreak.com
 
Old 05-12-2011, 09:23 AM
Markus Neviadomski
 
Default Help About Squeeze

Am 12.05.2011 10:58, schrieb Stan Hoeppner:
> On 5/12/2011 2:14 AM, Markus Neviadomski wrote:
>> Hi,
>>
>> Thats sound like a fake, sorry. A single-core CPU with a some piece of
>> RAM as file server for 200 users? No...
>>
>> Buy a new hardware with 4 or 6 cores, 16GB RAM and a raid controller and
>> some disks. Then copy your data from the old system on the new raid
>> system and everything is fine.
>
> You didn't answer the OP's question.

I did. The OP should get a hardware without errors (its the best time to
upgrade the hardware to a new one, but any old one is also ok) and do a
correct data recovery from his old disks to a new raid system. He has
200 users behind his server. Nobody could do any work until the OP is
testing some BIOS settings or system configs without sleep mode,
recovering the file system and so on. Because of his question, the OP is
not very confident dealing with hard- and software, a new setup is the
best way with the fewest possible errors.


> And your upgrade recommendation is complete overkill. A modern ~2Ghz
> Sempron can easily saturate a GbE pipe without using jumbo frames.
> 1GB of RAM is plenty for serving 200 office environment Samba
> clients--512MB would even be sufficient. If he's lacking performance
> it's due to insufficient head seeks bandwidth. He didn't mention a
> capacity shortage and he currently has ~250GB of disk. A single 250GB
> SSD would solve that problem instantly for $400-600 USD with an MLC
> drive. Two mirrored would be better, so $400-1200 USD. This would
> yield 50,000 seeks/second vs the 300 he has now, and ~250MB/s
> bandwidth, vs the ~160MB/s he currently has.

Never use a single drive for important data. Dont tell this idea to
other people please!

>
>


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4DCBA71C.2060804@dieitexperten.de">http://lists.debian.org/4DCBA71C.2060804@dieitexperten.de
 
Old 05-12-2011, 09:34 AM
Markus Neviadomski
 
Default Help About Squeeze

Am 12.05.2011 11:20, schrieb Stan Hoeppner:
> On 5/12/2011 3:29 AM, Markus Neviadomski wrote:
>
>> And the costs of hardware could be igored, if you have to recover broken
>> or lost data w/o raid or backup!
>
> This typically holds true in the US where the total cost of labor is
> far greater than hardware. This is definitely not the case in
> 'developing' countries, where the acquisition cost of a single tier 1
> server may very well be greater than an SA's yearly salary.
>
I know about that. But lost data couldnt be recovered, only rewritten.
There admin costs will get zero, if 200 users has to be rewrite all
there documents. The admin will need a grave!

> ~$20k can buy you a 2 socket 24 core AMD Magny Cours HP server with
> 32GB RAM, quad GbE ports, a 10 GbE PCIe x4/x8 NIC, LSI's top of the
> line PCIe x8 RAID HBA with 1GB BBWC and 2 SFF8088 SAS ports, two LSI
> 24 drive 2.5" chainable SAS enclosures w/ internal expanders and 48
> SAS drives of 300GB capacity and 15k spindle speed.

You are right...but, read my first post. I wrote about 4-6 cores. This
is a typical sandy bridge system with core i7. No real server hardware,
but fast enough for the OP, when his Sempron is working until today!
16GB Ram are possible and cheap enough for development countrys servers.

>
> The cost of this system, in many parts of the world, may be double (or
> more) the yearly salary of the SA managing it. In the US this
> system's price tag will equal about 1/4 to 1/6th the SA's total yearly
> compensation package, depending on the SA's state/city.
>


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4DCBA9A6.9030504@dieitexperten.de">http://lists.debian.org/4DCBA9A6.9030504@dieitexperten.de
 
Old 05-12-2011, 09:41 AM
Stan Hoeppner
 
Default Help About Squeeze

On 5/12/2011 4:23 AM, Markus Neviadomski wrote:

Am 12.05.2011 10:58, schrieb Stan Hoeppner:



a new setup is the
best way with the fewest possible errors.


It's becoming very clear you've not been in the SA game very long...


Two mirrored would be better, so $400-1200 USD. This would
yield 50,000 seeks/second vs the 300 he has now, and ~250MB/s
bandwidth, vs the ~160MB/s he currently has.



Never use a single drive for important data. Dont tell this idea to
other people please!


Apparently you don't read thoroughly either. Note my words "Two
mirrored" above.


--
Stan


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Archive: 4DCBAB2E.9060603@hardwarefreak.com">http://lists.debian.org/4DCBAB2E.9060603@hardwarefreak.com
 

Thread Tools




All times are GMT. The time now is 12:30 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org