[<prev] [next>] [day] [month] [year] [list]
Message-ID: <f1dc79790612172325n54e1e654u76c31120d4153cf5@mail.gmail.com>
Date: Mon, 18 Dec 2006 10:25:01 +0300
From: "Nick Olkin" <nick.olkin@...il.com>
To: linux-kernel@...r.kernel.org
Cc: nick.olkin@...il.com
Subject: Q: pages_min determination
Dear all!
I've a question regarding the work of the Virtual Memory Manager.
Unfortunately reading the "Understanding The Linux Virtual Memory
Manager" (Mel Gorman) didn't help much.
In general the question is how to determine on the working machine
what the number of the pages_min is?
I have a SunFire v20z running the Red Hat Enterprise Linux AS release
3 (Taroon Update 7) 2.4.21-40.ELsmp.
To be honest this question is not of very high urgency and it does not
imply high practical importance. The interest was stimulated by the
observing the VM activities on the machine. The server has 4G of
physical RAM and 4G of swap. A Java App is running on this host. It
works quite stable, but any time I check the memory state I see that
the free memory is quite low, and the swap is not used at all. It does
not worry me as far as I don't see any lack of memory messages, but I
wonder how the zone balancing occurs, how does it determine the way of
freeing pages and how can make the system to start swaping (just for
testing of cause). I tried to load my system a bit more (started
different applications to consume the memory), but as you all guessed,
all I've got is slightly lessened free mem and some reclamation
between cached/used/free.
I thought the value of min_free_kbytes would help me, but I didn't
find it on my system:
ls /proc/sys/vm/
bdflush
dcache_priority
hugetlb_pool
inactive_clean_percent
kscand_work_percent
kswapd
max_map_count
max-readahead
min-readahead
oom-kill
overcommit_memory
overcommit_ratio
pagecache
page-cluster
pagetable_cache
skip_mapped_pages
stack_defer_threshold
The sample output of free –m
total used free shared buffers cached
Mem: 3955 3927 28 0 65 3611
-/+ buffers/cache: 250 3705
Swap: 4094 0 4094
I will greatly appreciate any assistance although I understand that my
question does not suite this mailing-list entirely.
Thanks a lot!
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists