Do you have a verifiable (as in from a knowledgeable source) reference
for this? - it goes against a lot of what I found googling a year ago
where swap size was dependent on CPU architecture (i.e.,
zeon/opteron/athlon etc), not 32/64bit.)
e.g., see How large can my swap space be? at
On 1/19/2011 12:07 AM, William Kenworthy wrote:
Do you have a verifiable (as in from a knowledgeable source) reference
for this? - it goes against a lot of what I found googling a year ago
where swap size was dependent on CPU architecture (i.e.,
zeon/opteron/athlon etc), not 32/64bit.)
You
On Tuesday 18 January 2011 04:42:38 William Kenworthy wrote:
On Mon, 2011-01-17 at 16:23 -0800, Grant wrote:
And for those with gigabytes of swap, keep in mind that the majority of
processors can only access up to 32 x 2G swapfiles under linux, so 4G is
only going to be half used.
Unless
Apparently, though unproven, at 09:27 on Tuesday 18 January 2011, Mick did
opine thusly:
BTW, it used to be that the kernel would not (easily?) access more than
128M of swap for some reason and multiple 128M swap partitions were more
efficient than a single larger space, but this has
On Tue, 2011-01-18 at 07:27 +, Mick wrote:
On Tuesday 18 January 2011 04:42:38 William Kenworthy wrote:
On Mon, 2011-01-17 at 16:23 -0800, Grant wrote:
And for those with gigabytes of swap, keep in mind that the majority of
processors can only access up to 32 x 2G swapfiles under
On 1/17/2011 8:42 PM, William Kenworthy wrote:
No swap contains pages from memory that have not been accessed for
awhile so they can be stored elsewhere freeing ram for actual active
pages. When they need to be accessed, they have to be swapped back in,
and often something swapped back out to
Apparently, though unproven, at 03:21 on Monday 17 January 2011, William
Kenworthy did opine thusly:
A
modern desktop that swaps is unusable - enormous amounts of data has to
be pulled back in from the drive. A web server that swaps is already
thrashing so you always want to avoid that.
Apparently, though unproven, at 03:39 on Monday 17 January 2011, William
Kenworthy did opine thusly:
On Sun, 2011-01-16 at 17:26 -0800, Mark Knecht wrote:
On Sun, Jan 16, 2011 at 5:13 PM, William Kenworthy bi...@iinet.net.au
wrote:
On Sun, 2011-01-16 at 14:41 -0800, Grant wrote:
...
On Mon, 2011-01-17 at 10:07 +0200, Alan McKinnon wrote:
Apparently, though unproven, at 03:39 on Monday 17 January 2011, William
Kenworthy did opine thusly:
On Sun, 2011-01-16 at 17:26 -0800, Mark Knecht wrote:
On Sun, Jan 16, 2011 at 5:13 PM, William Kenworthy bi...@iinet.net.au
Apparently, though unproven, at 10:18 on Monday 17 January 2011, William
Kenworthy did opine thusly:
On Mon, 2011-01-17 at 10:07 +0200, Alan McKinnon wrote:
Apparently, though unproven, at 03:39 on Monday 17 January 2011, William
Kenworthy did opine thusly:
On Sun, 2011-01-16 at 17:26
On Mon, 17 Jan 2011 10:07:45 +0200, Alan McKinnon wrote:
I have a diskless 3GB ram atom system (mythtv frontend) and I have to
arrange swap over nbd for gcc and glibc emerges - others just get very
slow when getting to limits, or get flaky unless -j1 is used. Havent
tried OO on it yet :)
On 17/1/2011, at 8:07am, Alan McKinnon wrote:
...
I'M flabbergasted. 3G is really a gigantic amount of memory and yet the
machine still runs out of the stuff?
Something is seriously wrong somewhere when code does this. I know memory is
cheap and all, but still ... that's just excessive
On Mon, 2011-01-17 at 09:22 +, Neil Bothwick wrote:
On Mon, 17 Jan 2011 10:07:45 +0200, Alan McKinnon wrote:
I have a diskless 3GB ram atom system (mythtv frontend) and I have to
arrange swap over nbd for gcc and glibc emerges - others just get very
slow when getting to limits, or
On Mon, 17 Jan 2011 19:40:15 +0800, William Kenworthy wrote:
If it's diskless, where are /tmp and /var/tmp mounted? If they use
tmpfs the memory usage is understandable. If they use NFS the
emerges must be unbearably slow.
For normal usage, they are in tmpfs along with portage but I mount
On Mon, 2011-01-17 at 12:38 +, Neil Bothwick wrote:
On Mon, 17 Jan 2011 19:40:15 +0800, William Kenworthy wrote:
If it's diskless, where are /tmp and /var/tmp mounted? If they use
tmpfs the memory usage is understandable. If they use NFS the
emerges must be unbearably slow.
For
On 1/17/2011 12:29 AM, Alan McKinnon wrote:
Not so much :-)
I too have db servers with 96G of ram. 5 of them, so I'm current. I'm just
gobsmacked that a desktop needs 3G to build a compiler and system libs. It's
consuming 2G to do that, I'll bet that 1.75G of that is pure wastage.
Much like
On Monday 17 January 2011 22:45:39 kashani wrote:
On 1/17/2011 12:29 AM, Alan McKinnon wrote:
Not so much :-)
I too have db servers with 96G of ram. 5 of them, so I'm current. I'm
just gobsmacked that a desktop needs 3G to build a compiler and system
libs. It's consuming 2G to do that,
On Monday 17 January 2011 12:38:56 Neil Bothwick wrote:
On Mon, 17 Jan 2011 19:40:15 +0800, William Kenworthy wrote:
If it's diskless, where are /tmp and /var/tmp mounted? If they use
tmpfs the memory usage is understandable. If they use NFS the
emerges must be unbearably slow.
For
I think the idea is never use swap if possible, but in a case where
you don't have swap space or run out of swap space I think it's still
possible to lose data.
Isn't swap just an extension of system memory? Isn't adding 4GB of
memory just as effective at preventing out-of-memory as
On Mon, 17 Jan 2011 23:44:59 +, Mick wrote:
So root is on another (more powerful?) machine and mounted over NFS?
Why not chroot into the root on the host machine and run the emerge
there?
Is there a howto for this somewhere please?
It's just the same as if you'd booted from a live
On 1/17/2011 4:23 PM, Grant wrote:
I think the idea is never use swap if possible, but in a case where
you don't have swap space or run out of swap space I think it's still
possible to lose data.
Isn't swap just an extension of system memory? Isn't adding 4GB of
memory just as effective at
On Mon, Jan 17, 2011 at 4:23 PM, Grant emailgr...@gmail.com wrote:
I think the idea is never use swap if possible, but in a case where
you don't have swap space or run out of swap space I think it's still
possible to lose data.
Isn't swap just an extension of system memory? Isn't adding 4GB
Grant wrote:
I think the idea is never use swap if possible, but in a case where
you don't have swap space or run out of swap space I think it's still
possible to lose data.
Isn't swap just an extension of system memory? Isn't adding 4GB of
memory just as effective at preventing
On Mon, 2011-01-17 at 16:23 -0800, Grant wrote:
I think the idea is never use swap if possible, but in a case where
you don't have swap space or run out of swap space I think it's still
possible to lose data.
Isn't swap just an extension of system memory? Isn't adding 4GB of
memory just
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back on? I have 3GB RAM and MAKEOPTS=-j1.
- Grant
On Sunday 16 January 2011 11:25:06 Grant wrote:
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back on? I have 3GB RAM
On 01/16/2011 08:25 PM, Grant wrote:
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back on? I have 3GB RAM and
On Sun, Jan 16, 2011 at 11:25 AM, Grant emailgr...@gmail.com wrote:
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back on? I have 3GB RAM and MAKEOPTS=-j1.
- Grant
The near freeze is
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back on? I have 3GB RAM and MAKEOPTS=-j1.
- Grant
As Volker says,
On Sun, Jan 16, 2011 at 2:41 PM, Grant emailgr...@gmail.com wrote:
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back
Apparently, though unproven, at 02:15 on Monday 17 January 2011, Mark Knecht
did opine thusly:
[snip]
As Volker says, don't turn swap off. Make it small if you must, but
keep some around. It's just disk space.
I thought swap was no longer necessary on a machine with sufficient
On Sun, 2011-01-16 at 14:41 -0800, Grant wrote:
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there anything I can do to prevent this
besides turning swap back on? I have 3GB
On Mon, 2011-01-17 at 03:08 +0200, Alan McKinnon wrote:
Apparently, though unproven, at 02:15 on Monday 17 January 2011, Mark Knecht
did opine thusly:
[snip]
As Volker says, don't turn swap off. Make it small if you must, but
keep some around. It's just disk space.
I thought
On Sun, Jan 16, 2011 at 5:13 PM, William Kenworthy bi...@iinet.net.au wrote:
On Sun, 2011-01-16 at 14:41 -0800, Grant wrote:
I've been running without swap for quite a while, but my system goes
into a near freeze whenever I undertake a large emerge such as
chromium or openoffice. Is there
On Sun, 16 Jan 2011 14:41:24 -0800, Grant wrote:
'onice -c 3 emerge -DuN world' ended up working great.
Or you can set PORTAGE_IONICE_COMMAND in make.conf to make it a default.
--
Neil Bothwick
Vuja De: the feeling that you've never been here before.
signature.asc
Description: PGP
On Sun, 2011-01-16 at 17:26 -0800, Mark Knecht wrote:
On Sun, Jan 16, 2011 at 5:13 PM, William Kenworthy bi...@iinet.net.au wrote:
On Sun, 2011-01-16 at 14:41 -0800, Grant wrote:
...
I think that's well worded. He has insufficient memory when emerging.
If he's really running short of DRAM
37 matches
Mail list logo
