From 1be510f9529cb082f802408b472a77d074b394c0 Mon Sep 17 00:00:00 2001 From: Nicolas Vigier Date: Sun, 14 Apr 2013 13:46:12 +0000 Subject: Add zarb MLs html archives --- zarb-ml/mageia-sysadm/2010-November/000132.html | 101 ++++++++++++++++++++++++ 1 file changed, 101 insertions(+) create mode 100644 zarb-ml/mageia-sysadm/2010-November/000132.html (limited to 'zarb-ml/mageia-sysadm/2010-November/000132.html') diff --git a/zarb-ml/mageia-sysadm/2010-November/000132.html b/zarb-ml/mageia-sysadm/2010-November/000132.html new file mode 100644 index 000000000..9de72e553 --- /dev/null +++ b/zarb-ml/mageia-sysadm/2010-November/000132.html @@ -0,0 +1,101 @@ + + + + [Mageia-sysadm] 2010.1 kernel with the vmscan fix + + + + + + + + + +

[Mageia-sysadm] 2010.1 kernel with the vmscan fix

+ Buchan Milne + bgmilne at mandriva.org +
+ Mon Nov 1 11:47:51 CET 2010 +

+
+ +
On Sunday, 24 October 2010 07:28:55 Thomas Backlund wrote:
+> Hi,
+> there is now a kernel-2.6.33.7-2.1mnb for 2010.1 at:
+> 
+> http://tmb.mine.nu/Mandriva/2010.1/
+> http://tmb2.mine.nu/Mandriva/2010.1/
+> 
+> With the fix:
+> - vmscan: raise the bar to PAGEOUT_IO_SYNC stalls
+>    (Fixes "system goes unresponsive under memory pressure and lots of
+>     dirty/writeback pages" bug. (http://lkml.org/lkml/2010/4/4/86))
+> 
+> 
+> This was also suggested by blino to help the BS
+
+This is definitely a huge improvement:
+
+[root at tiger ~]# uname -r
+2.6.33.7-desktop-2.1mnb
+[root at tiger ~]# time dd if=/media/smartdrive/isos/mandriva/mandriva-linux-
+free-2010-spring-i586/mandriva-linux-free-2010-spring-i586.iso 
+of=/dev/sdc;uptime
+9107456+0 records in
+9107456+0 records out
+4663017472 bytes (4.7 GB) copied, 7075.81 s, 659 kB/s
+4.83user 60.63system 1:57:55elapsed 0%CPU (0avgtext+0avgdata 3216maxresident)k
+18186800inputs+9107456outputs (1major+260minor)pagefaults 0swaps
+ 11:38:07 up 17:35, 21 users,  load average: 4.57, 5.00, 5.23
+
+While this was running, the system was quite responsive, in fact I didn't 
+notice that it was under load, whereas before, with dd'ing an ISO from a USB 
+hard drive to a slow USB flash disk, load average was higher, and my system 
+was almost unusable (even just switching between running applications was 
+tedious). Unfortunately, due to a 20-month-old having learned how to operate 
+switches, I booted into this kernel without notice, so I didn't validate my 
+previous experience on the previous kernel with exactly the same operation, 
+but I am quite sure I have experienced it before on 2.6.33.x.
+
+Regards,
+Buchan
+
+ + + + + + + + +
+

+ +
+More information about the Mageia-sysadm +mailing list
+ -- cgit v1.2.1