* Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
@ 2005-06-27 9:18 Ewerton Romulo
2005-06-27 14:20 ` Jan Hudec
0 siblings, 1 reply; 12+ messages in thread
From: Ewerton Romulo @ 2005-06-27 9:18 UTC (permalink / raw)
To: linux-fsdevel
I am building a very small linux system to boot off an IDE-HD module
and run from a ramdisk.
He´s working well and have no problem funtion like that. I´m using
linux kernel 2.4.30!
But some times I need to mount a partion on HD and use a directory to
save many files in it!
I mount it, create a directory there but ... when I try to use the HD
mount with the directory, save many files in it, the memory go up .. up
... and the linux stop working!!!
If I umount the partion everything return to function well.
When I mount the partion again and begin to put files in it ... the
problem return!
I try to mount partion using ext2 and ext3 ... the problem is the
same!!!
I believe that has something not working well with how the ramdisk or
initrd use the HD mounted!!!
I apreciate any help ...
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-27 9:18 Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow! Ewerton Romulo
@ 2005-06-27 14:20 ` Jan Hudec
2005-06-28 18:06 ` Ewerton Romulo
0 siblings, 1 reply; 12+ messages in thread
From: Jan Hudec @ 2005-06-27 14:20 UTC (permalink / raw)
To: Ewerton Romulo; +Cc: linux-fsdevel
[-- Attachment #1: Type: text/plain, Size: 1312 bytes --]
On Mon, Jun 27, 2005 at 06:18:11 -0300, Ewerton Romulo wrote:
> I am building a very small linux system to boot off an IDE-HD module
> and run from a ramdisk.
Ramdisk *OR* tmpfs? They are completely different things.
> He´s working well and have no problem funtion like that. I´m using
> linux kernel 2.4.30!
> But some times I need to mount a partion on HD and use a directory to
> save many files in it!
How large are those "many files"? Where do the data come from?
> I mount it, create a directory there but ... when I try to use the HD
> mount with the directory, save many files in it, the memory go up .. up
> ... and the linux stop working!!!o
> If I umount the partion everything return to function well.
How do you umount it, when it stopped working?!
Anyway, could you provide contents of /proc/meminfo, /proc/slabinfo,
/proc/mounts and /proc/swaps when the bad things are happening?
> When I mount the partion again and begin to put files in it ... the
> problem return!
> I try to mount partion using ext2 and ext3 ... the problem is the
> same!!!
I'd be surprised if there was a difference. The filesystems are quite
similar.
-------------------------------------------------------------------------------
Jan 'Bulb' Hudec <bulb@ucw.cz>
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-27 14:20 ` Jan Hudec
@ 2005-06-28 18:06 ` Ewerton Romulo
2005-06-28 20:24 ` Jan Hudec
2005-06-28 21:42 ` Bryan Henderson
0 siblings, 2 replies; 12+ messages in thread
From: Ewerton Romulo @ 2005-06-28 18:06 UTC (permalink / raw)
To: Jan Hudec; +Cc: linux-fsdevel
Thanks for answer Jan
> > I am building a very small linux system to boot off an IDE-HD module
> > and run from a ramdisk.
>
> Ramdisk *OR* tmpfs? They are completely different things.
Sorry. I can say that is a ramdisk formatted to use tmpfs
filesystem! If I need to select one, so can I say that is a tmpfs!
> > He´s working well and have no problem funtion like that. I´m using
> > linux kernel 2.4.30!
> > But some times I need to mount a partion on HD and use a directory to
> > save many files in it!
>
> How large are those "many files"? Where do the data come from?
>
Any files!! Any size!!! From any where!!! If I´d like to save a .txt
file ... or download a .tgz file and save in it .. a gif figure to
using later ... just a jpeg ... a HTML file ... a big movie file ..
a mp3 file ... just a .bin ... or linux module ... download from ftp
... any size ... any file ... from any where!! If I mount a partion on
HD and put things in it ... consumes memory (RAM) ... and problem
appear!!!
Not important if the file is big or not, if I put things there I had
this problem!!
Suppose that I had 100MB in a partion mounted HD !!! If I put many
3KB txt in there ... I think that I´ll not problem with memory!!! But
it´s consumes my memory!!!
> > I mount it, create a directory there but ... when I try to use the HD
> > mount with the directory, save many files in it, the memory go up .. up
> > ... and the linux stop working!!!o
> > If I umount the partion everything return to function well.
>
> How do you umount it, when it stopped working?!
>
No! I umount before it´s stopped!!! Before memory go 100% .. And
things back to normal!
> Anyway, could you provide contents of /proc/meminfo, /proc/slabinfo,
> /proc/mounts and /proc/swaps when the bad things are happening?
I´ll take a look in it and post here!
>
> > When I mount the partion again and begin to put files in it ... the
> > problem return!
> > I try to mount partion using ext2 and ext3 ... the problem is the
> > same!!!
>
> I'd be surprised if there was a difference. The filesystems are quite
> similar.
In this case I only use it to see if I had different result! I try
to mount a fat partion too!! Same result!!!
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-28 18:06 ` Ewerton Romulo
@ 2005-06-28 20:24 ` Jan Hudec
2005-06-29 4:41 ` Ewerton Romulo
2005-06-28 21:42 ` Bryan Henderson
1 sibling, 1 reply; 12+ messages in thread
From: Jan Hudec @ 2005-06-28 20:24 UTC (permalink / raw)
To: Ewerton Romulo; +Cc: linux-fsdevel
[-- Attachment #1: Type: text/plain, Size: 1531 bytes --]
On Tue, Jun 28, 2005 at 15:06:55 -0300, Ewerton Romulo wrote:
> Thanks for answer Jan
>
> > > I am building a very small linux system to boot off an IDE-HD module
> > > and run from a ramdisk.
> >
> > Ramdisk *OR* tmpfs? They are completely different things.
>
> Sorry. I can say that is a ramdisk formatted to use tmpfs
> filesystem! If I need to select one, so can I say that is a tmpfs!
Sorry, it isn't. Nothing can be formated to use tmpfs, because tmpfs is
no-device filesystem. Tmpfs lives in RAM, but directly, without
a ramdisk. On the other hand ramdisk emulates a block device with memory
and has to be formated with some block device filesystem -- which tmpfs
isn't.
[skipped lot of exclamation marks with intersparsed small bits of text]
> > Anyway, could you provide contents of /proc/meminfo, /proc/slabinfo,
> > /proc/mounts and /proc/swaps when the bad things are happening?
>
> I´ll take a look in it and post here!
Ok. _when_ you post it, there will be something to talk about.
Oh, include also exact version of kernel you are using (including from
which distribution or whether it's official kernel), contents of
/proc/modules and if it's a kernel you compiled it's config.
> In this case I only use it to see if I had different result! I try
> to mount a fat partion too!! Same result!!!
That mostly rules out the filesystem driver. Ok.
-------------------------------------------------------------------------------
Jan 'Bulb' Hudec <bulb@ucw.cz>
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-28 18:06 ` Ewerton Romulo
2005-06-28 20:24 ` Jan Hudec
@ 2005-06-28 21:42 ` Bryan Henderson
1 sibling, 0 replies; 12+ messages in thread
From: Bryan Henderson @ 2005-06-28 21:42 UTC (permalink / raw)
To: Ewerton Romulo; +Cc: Jan Hudec, linux-fsdevel
> Sorry. I can say that is a ramdisk formatted to use tmpfs
>filesystem! If I need to select one, so can I say that is a tmpfs!
It's really easy to answer the question "what filesystem type is it?".
Just look in /proc/mounts. There's a line for each mounted filesystem,
and the 3rd word is the filesystem type.
There's a lot of your story that doesn't seem to be in conventional
terminology, so I don't really understand it.
But I'm starting to get a good idea of what you're trying to say. My
guess is you have a filesystem of some kind on a ramdisk. As Jan says, we
know it isn't "tmpfs", since that filesystem type doesn't live on a disk.
But if you're using a ramdisk, I doubt it makes any difference how the
data on it is formatted (i.e. the filesystem type).
Ramdisks have a characteristic similar to what you describe: A brand new
ramdisk contains all zeroes and uses no memory. The more blocks you write
to a ramdisk, the more real memory it occupies, and the memory is not
reclaimable in any way by the memory manager and the amount is not limited
by anything but the ramdisk size. If your ramdisk size is bigger than the
amount of real memory you have (or just bigger than the amount you can
spare), you will eventually kill your system if you keep putting data on
the ramdisk.
Unmounting a filesystem that is on a ramdisk wouldn't get that memory
back, but if unmounting stops you from writing any more, that might help.
Destroying the ramdisk would get the memory back.
The way you fix that problem is by making sure you don't allocate more
memory to ramdisks than you have. When you create a ramdisk, you say how
big it is, and you should be able to spare a byte of real memory for every
byte of ramdisk. It's a manual memory management task.
Or you might forget about ramdisks and use tmpfs instead.
BTW, what's with the exclamation points (!)? I think periods (.) are more
appropriate, and definitely easier to read, here. Similarly, the ellipses
(...) look like they're meant to be commas (,).
--
Bryan Henderson IBM Almaden Research Center
San Jose CA Filesystems
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-28 20:24 ` Jan Hudec
@ 2005-06-29 4:41 ` Ewerton Romulo
2005-06-29 8:41 ` Jan Hudec
2005-06-29 16:28 ` Bryan Henderson
0 siblings, 2 replies; 12+ messages in thread
From: Ewerton Romulo @ 2005-06-29 4:41 UTC (permalink / raw)
To: Jan Hudec; +Cc: linux-fsdevel
OK!
/proc/modules
via-rhine 10328 1
8139too 11996 1
mii 1868 0 [via-rhine 8139too]
sch_ingress 1116 0 (unused)
sch_sfq 2816 321
sch_htb 16896 2
cls_u32 3960 108
cls_fw 2004 1
ip_nat_quake3 1544 0 (unused)
ip_nat_mms 2352 0 (unused)
ip_nat_h323 2020 0 (unused)
ip_nat_amanda 708 0 (unused)
ip_nat_irc 1648 0 (unused)
ip_nat_ftp 2096 0 (unused)
ip_conntrack_quake3 1592 1
ip_conntrack_mms 2480 1
ip_conntrack_h323 1832 1
ip_conntrack_egg 2024 0 (unused)
ip_conntrack_amanda 968 1
ip_conntrack_irc 2344 1
ip_conntrack_ftp 3120 1
softdog 1316 1
My /proc/mount is:
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/dev/hda1 /test vfat rw 0 0
The tmpfs is my root that is descompression from a .tgz file in the
beginning of boot.
I use kernel 2.4.30 oficial, no especific linux distribution!
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-29 4:41 ` Ewerton Romulo
@ 2005-06-29 8:41 ` Jan Hudec
2005-06-29 16:28 ` Bryan Henderson
1 sibling, 0 replies; 12+ messages in thread
From: Jan Hudec @ 2005-06-29 8:41 UTC (permalink / raw)
To: Ewerton Romulo; +Cc: linux-fsdevel
[-- Attachment #1: Type: text/plain, Size: 1320 bytes --]
On Wed, Jun 29, 2005 at 01:41:13 -0300, Ewerton Romulo wrote:
> OK!
>
> /proc/modules
> [snipped]
Only network stuff here. Since there is no problem when disk is not
used, I think we can rule out network stuff.
> My /proc/mount is:
> rootfs / rootfs rw 0 0
> /dev/root / tmpfs rw 0 0
So you are using tmpfs, *not* ramdisk. Tmpfs lives in memory and
occupies amout of memory equal to amount of data stored. Unlike ramdisk,
memory is reclaimed when you remove files from tmpfs and when you umount
it.
> /proc /proc proc rw 0 0
> devpts /dev/pts devpts rw 0 0
> /dev/hda1 /test vfat rw 0 0
Ok, got the picture.
> I use kernel 2.4.30 oficial, no especific linux distribution!
You should have included your config too, since you are using a custom
one. We need to know what drivers you have compiled in your kernel. It's
called .config and located in the build directory of the kernel.
And of course, you did not include the most important stuff --
/proc/meminfo, /proc/slabinfo, /proc/diskstats and /proc/vmstats from
time when it exhibits the problem. Also you forgot about /proc/swaps (it
may be empty or not even exist, but you should at least say so).
-------------------------------------------------------------------------------
Jan 'Bulb' Hudec <bulb@ucw.cz>
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-29 4:41 ` Ewerton Romulo
2005-06-29 8:41 ` Jan Hudec
@ 2005-06-29 16:28 ` Bryan Henderson
2005-06-29 18:36 ` Ewerton Romulo
2005-06-29 18:46 ` Ewerton Romulo
1 sibling, 2 replies; 12+ messages in thread
From: Bryan Henderson @ 2005-06-29 16:28 UTC (permalink / raw)
To: Ewerton Romulo; +Cc: Jan Hudec, linux-fsdevel
Your subject line says "use a mounted HD for files." Does this mean
you're putting these files into a filesystem that lives on a real,
conventional disk device?
In your /proc mounts, I see one such filesystem mounted:
/dev/hda1 /test vfat rw 0 0
You also have a tmpfs filesystem mounted, which you mentioned a few times,
but I believe all you really said about it is that it's your root
filesystem. Or is this also the one in which you're stuffing files when
the problem occurs?
/dev/root / tmpfs rw 0 0
Do you have reason to believe that the fact of having a tmpfs root
filesystem is related to the problem?
--
Bryan Henderson IBM Almaden Research Center
San Jose CA Filesystems
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-29 16:28 ` Bryan Henderson
@ 2005-06-29 18:36 ` Ewerton Romulo
2005-06-29 21:00 ` Bryan Henderson
2005-06-29 18:46 ` Ewerton Romulo
1 sibling, 1 reply; 12+ messages in thread
From: Ewerton Romulo @ 2005-06-29 18:36 UTC (permalink / raw)
To: Bryan Henderson; +Cc: Jan Hudec, linux-fsdevel
[-- Attachment #1: Type: text/plain, Size: 16255 bytes --]
OK! Let´s go!!!
Below I put the /proc/ files wanted ... before HD mount ... after HD
mount ... when the memory problem occur ... and ... then when I umount
HD!!! for better viee look txt attach!!!
The .config of the kernel is here:
http://www.rictec.com.br/claudio/coyote/kernel-2.4.30/config-224-ide
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 5816320 25636864 0 0 2674688
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 25036 kB
MemShared: 0 kB
Buffers: 0 kB
Cached: 2612 kB
SwapCached: 0 kB
Active: 1360 kB
Inactive: 1276 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 25036 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 3 11 336 1 1 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 8 24 160 1 1 1
arp_cache 3 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 189 198 176 9 9 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 1 59 64 1 1 1
mnt_cache 10 59 64 1 1 1
inode_cache 853 872 464 107 109 1
dentry_cache 850 875 112 25 25 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 100 560 96 5 14 1
mm_struct 13 27 144 1 1 1
vm_area_struct 162 240 80 4 5 1
fs_cache 12 113 32 1 1 1
files_cache 13 18 416 2 2 1
signal_act 14 18 1296 5 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 94 94 2048 47 47 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 21 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 115 120 256 8 8 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 88 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 262 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/proc/swaps
Filename Type Size Used Priority
After mounting ...
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 5771264 25681920 0 32768 2678784
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 25080 kB
MemShared: 0 kB
Buffers: 32 kB
Cached: 2616 kB
SwapCached: 0 kB
Active: 1388 kB
Inactive: 1280 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 25080 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 3 11 336 1 1 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 5 24 160 1 1 1
arp_cache 1 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 249 264 176 12 12 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 2 59 64 1 1 1
mnt_cache 11 59 64 1 1 1
inode_cache 868 872 464 109 109 1
dentry_cache 864 875 112 25 25 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 144 560 96 5 14 1
mm_struct 11 27 144 1 1 1
vm_area_struct 145 240 80 4 5 1
fs_cache 10 113 32 1 1 1
files_cache 11 18 416 2 2 1
signal_act 12 18 1296 4 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 100 100 2048 50 50 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 22 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 169 180 256 12 12 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 87 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 266 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/dev/hda1 /mnt vfat rw 0 0
/proc/swaps
Filename Type Size Used Priority
The problem occur!!!! Look memory free!!!!
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 29626368 1826816 0 77824 22487040
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 1784 kB
MemShared: 0 kB
Buffers: 76 kB
Cached: 21960 kB
SwapCached: 0 kB
Active: 2060 kB
Inactive: 19996 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 1784 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 11 11 336 1 1 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 3 24 160 1 1 1
arp_cache 1 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 334 352 176 16 16 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 2 59 64 1 1 1
mnt_cache 11 59 64 1 1 1
inode_cache 898 904 464 113 113 1
dentry_cache 894 910 112 26 26 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 38920 38920 96 973 973 1
mm_struct 11 27 144 1 1 1
vm_area_struct 145 240 80 4 5 1
fs_cache 10 113 32 1 1 1
files_cache 11 18 416 2 2 1
signal_act 12 18 1296 4 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 106 108 2048 53 54 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 22 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 247 255 256 17 17 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 87 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 270 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/dev/hda1 /mnt vfat rw 0 0
/proc/swaps
Filename Type Size Used Priority
HD umounted!!!! Look memory free!!!!
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 9744384 21708800 0 0 2678784
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 21200 kB
MemShared: 0 kB
Buffers: 0 kB
Cached: 2616 kB
SwapCached: 0 kB
Active: 1364 kB
Inactive: 1272 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 21200 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 2 22 336 1 2 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 3 24 160 1 1 1
arp_cache 1 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 337 352 176 16 16 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 1 59 64 1 1 1
mnt_cache 10 59 64 1 1 1
inode_cache 853 904 464 107 113 1
dentry_cache 850 910 112 25 26 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 100 38920 96 5 973 1
mm_struct 11 27 144 1 1 1
vm_area_struct 145 240 80 4 5 1
fs_cache 10 113 32 1 1 1
files_cache 11 18 416 2 2 1
signal_act 12 18 1296 4 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 106 110 2048 53 55 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 21 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 247 255 256 17 17 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 87 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 270 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/proc/swaps
Filename Type Size Used Priority
[-- Attachment #2: test tmpfs.txt --]
[-- Type: text/plain, Size: 16211 bytes --]
Below I put the /proc/ files wanted ... before HD mount ... after HD mount ... when the memory problem occur ... and ... then when I umount HD!!!
The .config of the kernel is here:
http://www.rictec.com.br/claudio/coyote/kernel-2.4.30/config-224-ide
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 5816320 25636864 0 0 2674688
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 25036 kB
MemShared: 0 kB
Buffers: 0 kB
Cached: 2612 kB
SwapCached: 0 kB
Active: 1360 kB
Inactive: 1276 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 25036 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 3 11 336 1 1 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 8 24 160 1 1 1
arp_cache 3 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 189 198 176 9 9 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 1 59 64 1 1 1
mnt_cache 10 59 64 1 1 1
inode_cache 853 872 464 107 109 1
dentry_cache 850 875 112 25 25 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 100 560 96 5 14 1
mm_struct 13 27 144 1 1 1
vm_area_struct 162 240 80 4 5 1
fs_cache 12 113 32 1 1 1
files_cache 13 18 416 2 2 1
signal_act 14 18 1296 5 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 94 94 2048 47 47 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 21 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 115 120 256 8 8 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 88 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 262 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/proc/swaps
Filename Type Size Used Priority
After mounting ...
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 5771264 25681920 0 32768 2678784
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 25080 kB
MemShared: 0 kB
Buffers: 32 kB
Cached: 2616 kB
SwapCached: 0 kB
Active: 1388 kB
Inactive: 1280 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 25080 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 3 11 336 1 1 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 5 24 160 1 1 1
arp_cache 1 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 249 264 176 12 12 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 2 59 64 1 1 1
mnt_cache 11 59 64 1 1 1
inode_cache 868 872 464 109 109 1
dentry_cache 864 875 112 25 25 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 144 560 96 5 14 1
mm_struct 11 27 144 1 1 1
vm_area_struct 145 240 80 4 5 1
fs_cache 10 113 32 1 1 1
files_cache 11 18 416 2 2 1
signal_act 12 18 1296 4 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 100 100 2048 50 50 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 22 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 169 180 256 12 12 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 87 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 266 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/dev/hda1 /mnt vfat rw 0 0
/proc/swaps
Filename Type Size Used Priority
The problem occur!!!! Look memory free!!!!
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 29626368 1826816 0 77824 22487040
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 1784 kB
MemShared: 0 kB
Buffers: 76 kB
Cached: 21960 kB
SwapCached: 0 kB
Active: 2060 kB
Inactive: 19996 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 1784 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 11 11 336 1 1 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 3 24 160 1 1 1
arp_cache 1 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 334 352 176 16 16 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 2 59 64 1 1 1
mnt_cache 11 59 64 1 1 1
inode_cache 898 904 464 113 113 1
dentry_cache 894 910 112 26 26 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 38920 38920 96 973 973 1
mm_struct 11 27 144 1 1 1
vm_area_struct 145 240 80 4 5 1
fs_cache 10 113 32 1 1 1
files_cache 11 18 416 2 2 1
signal_act 12 18 1296 4 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 106 108 2048 53 54 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 22 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 247 255 256 17 17 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 87 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 270 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/dev/hda1 /mnt vfat rw 0 0
/proc/swaps
Filename Type Size Used Priority
HD umounted!!!! Look memory free!!!!
/proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 31453184 9744384 21708800 0 0 2678784
Swap: 0 0 0
MemTotal: 30716 kB
MemFree: 21200 kB
MemShared: 0 kB
Buffers: 0 kB
Cached: 2616 kB
SwapCached: 0 kB
Active: 1364 kB
Inactive: 1272 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 30716 kB
LowFree: 21200 kB
SwapTotal: 0 kB
SwapFree: 0 kB
/proc/slabinfo
slabinfo - version: 1.1
kmem_cache 56 72 108 2 2 1
ip_conntrack 2 22 336 1 2 1
tcp_tw_bucket 0 40 96 0 1 1
tcp_bind_bucket 2 203 16 1 1 1
tcp_open_request 0 59 64 0 1 1
inet_peer_cache 0 0 48 0 0 1
ip_fib_hash 13 203 16 1 1 1
ip_dst_cache 3 24 160 1 1 1
arp_cache 1 40 96 1 1 1
blkdev_requests 1536 1560 96 39 39 1
dnotify_cache 0 0 20 0 0 1
file_lock_cache 0 0 92 0 0 1
fasync_cache 0 0 16 0 0 1
uid_cache 0 0 32 0 0 1
skbuff_head_cache 337 352 176 16 16 1
sock 18 27 864 2 3 2
sigqueue 0 29 132 0 1 1
kiobuf 0 0 64 0 0 1
cdev_cache 238 312 48 4 4 1
bdev_cache 1 59 64 1 1 1
mnt_cache 10 59 64 1 1 1
inode_cache 853 904 464 107 113 1
dentry_cache 850 910 112 25 26 1
filp 83 105 112 3 3 1
names_cache 0 2 4096 0 2 1
buffer_head 100 38920 96 5 973 1
mm_struct 11 27 144 1 1 1
vm_area_struct 145 240 80 4 5 1
fs_cache 10 113 32 1 1 1
files_cache 11 18 416 2 2 1
signal_act 12 18 1296 4 6 1
size-131072(DMA) 0 0 131072 0 0 32
size-131072 0 0 131072 0 0 32
size-65536(DMA) 0 0 65536 0 0 16
size-65536 0 0 65536 0 0 16
size-32768(DMA) 0 0 32768 0 0 8
size-32768 0 1 32768 0 1 8
size-16384(DMA) 0 0 16384 0 0 4
size-16384 0 0 16384 0 0 4
size-8192(DMA) 0 0 8192 0 0 2
size-8192 0 1 8192 0 1 2
size-4096(DMA) 0 0 4096 0 0 1
size-4096 4 4 4096 4 4 1
size-2048(DMA) 0 0 2048 0 0 1
size-2048 106 110 2048 53 55 1
size-1024(DMA) 0 0 1024 0 0 1
size-1024 21 24 1024 6 6 1
size-512(DMA) 0 0 512 0 0 1
size-512 28 32 512 4 4 1
size-256(DMA) 0 0 256 0 0 1
size-256 247 255 256 17 17 1
size-128(DMA) 0 0 128 0 0 1
size-128 521 540 128 18 18 1
size-64(DMA) 0 0 64 0 0 1
size-64 87 118 64 2 2 1
size-32(DMA) 0 0 32 0 0 1
size-32 270 339 32 3 3 1
/proc/mounts
rootfs / rootfs rw 0 0
/dev/root / tmpfs rw 0 0
/proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/proc/swaps
Filename Type Size Used Priority
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-29 16:28 ` Bryan Henderson
2005-06-29 18:36 ` Ewerton Romulo
@ 2005-06-29 18:46 ` Ewerton Romulo
1 sibling, 0 replies; 12+ messages in thread
From: Ewerton Romulo @ 2005-06-29 18:46 UTC (permalink / raw)
To: Bryan Henderson; +Cc: Jan Hudec, linux-fsdevel
> Your subject line says "use a mounted HD for files." Does this mean
> you're putting these files into a filesystem that lives on a real,
> conventional disk device?
>
> In your /proc mounts, I see one such filesystem mounted:
>
> /dev/hda1 /test vfat rw 0 0
Yes! That´s right!! The files is many .. that diferents sizes ..
so I can´t put it all in RAM!
> You also have a tmpfs filesystem mounted, which you mentioned a few times,
> but I believe all you really said about it is that it's your root
> filesystem. Or is this also the one in which you're stuffing files when
> the problem occurs?
>
> /dev/root / tmpfs rw 0 0
Well, my root filesystem is tmpfs, always ... only the new files
that I like to save, will be saved in a mount HD!
>
> Do you have reason to believe that the fact of having a tmpfs root
> filesystem is related to the problem?
Well, the problem only happen when I mount HD and begin to save files in it!
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-29 18:36 ` Ewerton Romulo
@ 2005-06-29 21:00 ` Bryan Henderson
2005-06-30 16:02 ` Jan Hudec
0 siblings, 1 reply; 12+ messages in thread
From: Bryan Henderson @ 2005-06-29 21:00 UTC (permalink / raw)
To: Ewerton Romulo; +Cc: Jan Hudec, linux-fsdevel
Is there some symptom besides MemFree being small? Does the system get
slow? Ridiculously slow? Does it hang?
MemFree probably is not what you think it is. There's a common
misconception that Linux memory is divided into used and free, and when
the free memory is gone, you're done. In fact, memory has varying degrees
of freeness. MemFree is the most free -- it's so free that it's wasted.
It contains no data of any use whatsoever. To reduce waste, Linux tries
to keep MemFree small. The numbers you have are normal.
The "Inactive" line is also quite free. This memory contains data that
might be useful again later, but probably not, and is recreatable from
another copy. So when Linux needs memory, it can simply throw out the
contents of this memory and use it. Your "inactive" line shows you have
plenty of memory available. It apparently contains the cached copy of all
those files you wrote.
Some inactive memory is easier to reuse than other. If it contains dirty
cache (buffered file writes), then Linux has to write the data to the disk
before it can use the memory. So things slow down to disk speed. I would
expect that to be the case in your scenario.
This stuff changes all the time, and with it the format of /proc/meminfo.
Unlike most of the /proc/meminfos I've seen, this version has nothing
about dirty pages in it, so it's hard to tell exactly where the system
sits. I wonder if this information is somewhere else? Anybody?
There's one thing I don't get: Where is the tmpfs data? With no swap
space and only a couple megabytes of page cache, I can't figure it out.
Surely the root filesystem is bigger than a few megabytes? Does someone
understand this?
--
Bryan Henderson IBM Almaden Research Center
San Jose CA Filesystems
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow!
2005-06-29 21:00 ` Bryan Henderson
@ 2005-06-30 16:02 ` Jan Hudec
0 siblings, 0 replies; 12+ messages in thread
From: Jan Hudec @ 2005-06-30 16:02 UTC (permalink / raw)
To: Bryan Henderson; +Cc: Ewerton Romulo, linux-fsdevel
[-- Attachment #1: Type: text/plain, Size: 1252 bytes --]
On Wed, Jun 29, 2005 at 14:00:56 -0700, Bryan Henderson wrote:
> Is there some symptom besides MemFree being small? Does the system get
> slow? Ridiculously slow? Does it hang?
>
> [...]
>
> This stuff changes all the time, and with it the format of /proc/meminfo.
> Unlike most of the /proc/meminfos I've seen, this version has nothing
> about dirty pages in it, so it's hard to tell exactly where the system
> sits. I wonder if this information is somewhere else? Anybody?
I have noticed there are quite many buffer heads, but I am not sure what
would be normal amount.
> There's one thing I don't get: Where is the tmpfs data? With no swap
> space and only a couple megabytes of page cache, I can't figure it out.
> Surely the root filesystem is bigger than a few megabytes? Does someone
> understand this?
I believe he talked about small system. So it actually can be just a few
megabytes. It obviously looks like setup for a router (with all the
iptables stuff loaded), which needs just a busybox, iptables to set it
up and perhaps ssh server so it can be reconfigured over network.
-------------------------------------------------------------------------------
Jan 'Bulb' Hudec <bulb@ucw.cz>
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
end of thread, other threads:[~2005-06-30 16:12 UTC | newest]
Thread overview: 12+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2005-06-27 9:18 Run linux in ramdisk(tmpfs) and use a mounted HD for files cause memory and CPU overflow! Ewerton Romulo
2005-06-27 14:20 ` Jan Hudec
2005-06-28 18:06 ` Ewerton Romulo
2005-06-28 20:24 ` Jan Hudec
2005-06-29 4:41 ` Ewerton Romulo
2005-06-29 8:41 ` Jan Hudec
2005-06-29 16:28 ` Bryan Henderson
2005-06-29 18:36 ` Ewerton Romulo
2005-06-29 21:00 ` Bryan Henderson
2005-06-30 16:02 ` Jan Hudec
2005-06-29 18:46 ` Ewerton Romulo
2005-06-28 21:42 ` Bryan Henderson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).