LinuxQuestions.org
Visit Jeremy's Blog.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Slackware
User Name
Password
Slackware This Forum is for the discussion of Slackware Linux.

Notices

Reply
 
Search this Thread
Old 01-03-2014, 02:08 PM   #16
metaschima
Senior Member
 
Registered: Dec 2013
Distribution: Slackware
Posts: 1,315

Rep: Reputation: Disabled

I use Slackware 14.1 on an i5-4670, 8 GB RAM, 1 TB SATA3 disk. I don't see any hardware-based reason for the slowdown, although I have heard that Seagate produces several types of 2 TB HDDs, some of which are slow. Here's a comment on my 1 TB drive, but it is about 2 TB drives:

Quote:
ok, so this drive is listed as the "Seagate ST2000DM001" and guess what; other than that it sports 2 Terabytes, it tells you nothing whatever about what drive you'll end up with, because Seagate has chosen to obscure and omit relevant Data between different builds with vastly different performance.

The short advice: Only purchase versions xxExxxxx [and possibly x24xxxxx - x24 is unverified info so far, see notes below] of the 2TB model. This uses 2 platters and 4 heads.
It performs 30% better than the version with 3 platters, which has an xxFxxxxx [or possibly x36xxxxx] designation. Avoid those!

You'll need to contact the seller and ask them to check the code on the drive. If they can't verify, don't buy it, better to get a drive from a different company, where its hopefully not a surprise game of what's in the box.

S - SU - Suzhou China
W - WU - Wuxi China
Z - TK - Korat Thailand

F = 3 platters with either 5 or 6 heads (bad 2TB drive or good 3TB drive)
E = 2 platters with 4 heads. (good 2TB drive)
D = 1 platter with 2 heads. (good 1TB drive)

Weight info received in a comment here, suggests that the 'good' 2-platter drive weighs 534 grams, while the 'bad' 3-platter drive weighs 624 grams.

Seagate used to embed the information about their drives in the model number, but now they obscured it, so they can pawn off whatever they want. Send a WxE model to Publications who test drives, and then ship the crappy WxF model to unsuspecting customers who may never realize they're not getting what they thought they were buying. This should really be illegal.
http://www.amazon.com/Seagate-Barrac...barracuda+1+TB

I'm not sure if this is the issue.

For the software side I have a custom kernel, but the only main difference is the 'deadline' I/O scheduler being the default, but you can test them out on-the-fly with the link I posted earlier.
 
Old 01-03-2014, 02:22 PM   #17
Petri Kaukasoina
Member
 
Registered: Mar 2007
Posts: 242

Rep: Reputation: 86
Quote:
Originally Posted by roderich View Post
16GB RAM

The kernel runs in 32-bit PAE mode, because for various reasons I did not (and still do not) want a full 64-bit system.
Try add a kernel boot parameter 'mem=3G'.
 
Old 01-03-2014, 03:18 PM   #18
chicken76
Member
 
Registered: Mar 2009
Distribution: Slackware
Posts: 77

Rep: Reputation: 1
Have you checked if the drive has the latest firmware?
While searching for a new drive to buy, I found a lot of threads about cheap Seagates in the 1-3 TB range having serious problems with early firmwares.
 
Old 01-03-2014, 03:50 PM   #19
syg00
LQ Veteran
 
Registered: Aug 2003
Location: Australia
Distribution: Lots ...
Posts: 12,350

Rep: Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041
Quote:
Originally Posted by roderich View Post
My current suspects go into the direction multiprocessing, memory management.
The kernel runs in 32-bit PAE mode, because for various reasons I did not (and still do not) want a full 64-bit system.
In which case you might be largely on your own - Linus said years ago he won't accept bug reports against PAE kernels on 64-bit hardware with large (> 8 Gig) memory.
Quote:
I will try to make some tests in a quiet hour with other Linux Live CDs and other kernel options (non-SMP, non-PAE, 64-bit) to see if these make any difference.
Good idea to narrow the field down.
 
Old 01-03-2014, 04:14 PM   #20
metaschima
Senior Member
 
Registered: Dec 2013
Distribution: Slackware
Posts: 1,315

Rep: Reputation: Disabled
Quote:
Originally Posted by chicken76 View Post
Have you checked if the drive has the latest firmware?
While searching for a new drive to buy, I found a lot of threads about cheap Seagates in the 1-3 TB range having serious problems with early firmwares.
That's a good idea. The 'smartctl -a' command should give you the info you need to see if there is an updated firmware:
https://apps1.seagate.com/downloads/...eCookie=en_US_

I just checked it with your info and there is an update:
http://knowledge.seagate.com/article...S/FAQ/223651en

Last edited by metaschima; 01-03-2014 at 04:16 PM.
 
Old 01-04-2014, 08:52 AM   #21
roderich
LQ Newbie
 
Registered: Dec 2013
Posts: 15

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by metaschima View Post
That's a good idea. The 'smartctl -a' command should give you the info you need to see if there is an updated firmware:
https://apps1.seagate.com/downloads/...eCookie=en_US_

I just checked it with your info and there is an update:
http://knowledge.seagate.com/article...S/FAQ/223651en
Thanks for the hint. Yes, the firmware available there seems to be newer than mine (CC4H vs. CC4G).
For the moment, however, I hesitate to make such a potentially harmful firmware update "just for fun",
first I will further explore the Linux kernel software track.
Remember that I observe the same symptoms also on several external USB disks
 
Old 01-04-2014, 09:05 AM   #22
roderich
LQ Newbie
 
Registered: Dec 2013
Posts: 15

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by syg00 View Post
In which case you might be largely on your own - Linus said years ago he won't accept bug reports against PAE kernels on 64-bit hardware with large (> 8 Gig) memory.
Hmm, that's interesting. This (PAE with 16GB RAM) could be a candidate for the "unique feature" of my system, which distinguishes it from all the others who do not have such a disk write speed problem.

In my general searches in the net I have meanwhile found one other discussion thread, where someone claims that a change from 32-bit PAE to 64-bit resolved such a problem.

See here: http://askubuntu.com/questions/33274...-ubuntu-system

Just out of "academic curiosity": Do you happen to have a reference to the current kernel developers' policy about this topic (something a bit more specific than "Linus said years ago"?

I'll make some further experiments in this direction.
 
Old 01-04-2014, 12:25 PM   #23
Petri Kaukasoina
Member
 
Registered: Mar 2007
Posts: 242

Rep: Reputation: 86
Quote:
Originally Posted by roderich View Post
Just out of "academic curiosity": Do you happen to have a reference to the current kernel developers' policy about this topic (something a bit more specific than "Linus said years ago"?
http://cl4ssic4l.wordpress.com/2011/05/24/linus-torvalds-about-pae/

http://yarchive.net/comp/linux/highmem.html

(These are old but I don't think anyone has had any interest in 32-bit kernels and gigabytes of memory for years.)

Last edited by Petri Kaukasoina; 01-04-2014 at 12:34 PM.
 
Old 01-04-2014, 12:59 PM   #24
Petri Kaukasoina
Member
 
Registered: Mar 2007
Posts: 242

Rep: Reputation: 86
This is less than a year old: https://lkml.org/lkml/2013/1/31/7
The whole thread is worth reading. The first message was https://lkml.org/lkml/2013/1/11/581
 
Old 01-05-2014, 05:50 AM   #25
roderich
LQ Newbie
 
Registered: Dec 2013
Posts: 15

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by Petri Kaukasoina View Post
This is less than a year old: https://lkml.org/lkml/2013/1/31/7
The whole thread is worth reading. The first message was https://lkml.org/lkml/2013/1/11/581
Thanks for this and the other links. Very interesting reads indeed.

What I read in this texts and what was new for me is that this PAE feature really creates a substantial overhead which surprisingly even gets larger the more memory you have. And what I observed with my disk is therefore probably not a hardware/software malfunction but a principal PAE problem.

I'm still preparing a test scenario where I can test various configurations on my machine under the same conditions.
I'll come back when I have new figures
 
Old 01-05-2014, 06:46 AM   #26
syg00
LQ Veteran
 
Registered: Aug 2003
Location: Australia
Distribution: Lots ...
Posts: 12,350

Rep: Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041Reputation: 1041
Just for completeness, read this - note that it's from 2007.

'nuff said.
 
Old 01-05-2014, 06:57 AM   #27
Petri Kaukasoina
Member
 
Registered: Mar 2007
Posts: 242

Rep: Reputation: 86
Quote:
Originally Posted by roderich View Post
I'm still preparing a test scenario where I can test various configurations on my machine under the same conditions.
Please try the 32-bit kernel also with "mem=880M" so you see it without highmem.
 
Old 01-05-2014, 08:02 AM   #28
roderich
LQ Newbie
 
Registered: Dec 2013
Posts: 15

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by roderich View Post

I'm still preparing a test scenario where I can test various configurations on my machine under the same conditions.
I'll come back when I have new figures
I *am* back, and now I am *totally confused*.
We had so many interesting theories ....
But now I have got the classical dentist effect: the pain disappears immediately when the dentist drill approaches.

I have made my systematic experiments with Slackware32 and Slackware64 Boot-DVDs with various memory limits, and I suddenly had write speeds of about 130 MB/s in *all* variations.
And after rebooting to my "normal" system it stayed that way !!!!!

So for the moment I can only conclude that it must have been some very special system situation (which just happened to occur a few days after my 14.1 system upgrade) which triggered the slowdown in the first place and disappeared again after some restarts. I will, nevertheless, keep an cautious eye on the issue.

Thanks, anyway, for the many hints and proposals.
 
Old 01-05-2014, 12:47 PM   #29
Richard Cranium
Senior Member
 
Registered: Apr 2009
Location: Carrollton, Texas
Distribution: Slackware64 14.1
Posts: 1,513

Rep: Reputation: 444Reputation: 444Reputation: 444Reputation: 444Reputation: 444
What happens when you power-cycle versus reboot?

One of your experiments may have set something on the drive that persists until power cycle. I only ask because a reboot isn't exactly the same thing as a power cycle.
 
Old 01-06-2014, 04:35 AM   #30
guanx
Senior Member
 
Registered: Dec 2008
Posts: 1,014

Rep: Reputation: 146Reputation: 146
Quote:
Originally Posted by roderich View Post
...
But now I have got the classical dentist effect: the pain disappears immediately when the dentist drill approaches.

I have made my systematic experiments with Slackware32 and Slackware64 Boot-DVDs with various memory limits, and I suddenly had write speeds of about 130 MB/s in *all* variations.
And after rebooting to my "normal" system it stayed that way !!!!!

...
Any mechanical resonance?

Loose SATA cable?

Failing disk (check with smartctl)?

Or (though unlikely from your previous posts) a degraded RAID being rebuilt?
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
How can external disk be mounted so a ordinary user can write to the disk Glenn D. Linux - Newbie 1 10-06-2012 02:32 AM
Sata disk write speed is very slow! eltonsky Debian 9 09-16-2010 06:07 AM
Extremely slow write on disk with high CPU usage aliojjati Linux - Newbie 1 09-04-2010 01:29 AM
slow disk reads: Slackware 10 kernel 2.6.7 nirvanix Slackware 7 06-29-2004 11:26 AM
slackware: slow disk access ? Menestrel Slackware 3 05-22-2004 05:46 PM


All times are GMT -5. The time now is 05:19 AM.

Main Menu
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
identi.ca: @linuxquestions
Facebook: linuxquestions Google+: linuxquestions
Open Source Consulting | Domain Registration