Re: [linux-lvm] User count, general question for all y'all

Re: [linux-lvm] User count, general question for all y'all

[Heinz Mauelshagen]

> 
> I haven't experience any I/O deadlocks, but I'm using IDE disks and a
> 2.0.36 kernel with ext2 fses.  Has anyone attempted using of the JFSes out
> there with LVM?  
> 

No, i didn't test with a JFS.

Only with reiserfs which worked o.k. so far.

Only the below mentioned restrictions applied.

BTW: i have not been able to test any performance boosts with reiserfs;
     reiserfsck is far to slow to be usefull


Heinz

P.S.: does somebody know about any Linux JFSes available for testing?


> At 07:22 AM 2/16/99 , you wrote:
> >
> >> 
> >> Are we all (on this list) actively using lvm, and in what
> >> configuration? How many of us are there?
> >> 
> >
> >I use it 8*)))
> >
> >> Also, has anyone encountered deadlocks during heavy IO?
> >
> >Yes, i did from the early 2.1.x times on.
> >The question is which Linux version you are using _and_
> >which aic7xxx version if any.
> >
> >It turned out to be a mass i/o lock which can be seen without LVM too.
> >Things turned out to be better in the later 2.1.x and in 2.2.x.
> >
> >A problem still stays which Stephen Tweedie and others are addressing right 
> >now:
> >with large quantity i/o 2.2.1 preferes allocation of buffer pages vs.
> >keeping process pages in memory making the system _very_ lazy ;*)
> >
> >
> >> I'm
> >> only using LVM at this point as a method of easy partitioning.
> >> No mirroring, etc. I suspect reiserfs is to blaim on my system,
> >> but I just thought I'd ask.
> >> 
> >
> >Mybe it's a big mkreiserfs forcing the metioned behaviour?
> >
> >Heinz
>
> 		Daniel Whicker  (heimdall@mail.org)
> 
> 


--

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Systemmanagement C/S                             Deutsche Telekom AG
                                                 Entwicklungszentrum Darmstadt
Heinz Mauelshagen                                Otto-Roehm-Strasse 71c
Senior Systems Engineer                          Postfach 10 05 41
                                                 64205 Darmstadt
mge@ez-darmstadt.telekom.de                      Germany
                                                 +49 6151 886-425
                                                          FAX-386
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: [linux-lvm] User count, general question for all y'all

[Dori Seliskar]

Heinz Mauelshagen wrote:

>
> P.S.: does somebody know about any Linux JFSes available for testing?
>

I'm working on a HA setup and every failover demands
fsck that  would be very nicely prevented by JFS.

I would be also happy to test it, but I believe first usable implementation
is/will be from Stephen Tweedie and he said that it won't be ready for testing
until some time this summer.


Regards,

Dori Seliskar

Re: [linux-lvm] User count, general question for all y'all

[Stephen C. Tweedie]

Hi,

On Tue, 16 Feb 1999 19:06:40 +0000, Dori Seliskar
<dori.seliskar@select-tech.si> said:

> I would be also happy to test it, but I believe first usable
> implementation is/will be from Stephen Tweedie and he said that it
> won't be ready for testing until some time this summer.

Current guesses are a month or two until usable demonstration code.

--Stephen

Re: [linux-lvm] User count, general question for all y'all

[Shawn Leas]

On Tue, 16 Feb 1999, Heinz Mauelshagen wrote:

> BTW: i have not been able to test any performance boosts with reiserfs;
>      reiserfsck is far to slow to be usefull

Reiser is slower unless you use notails,nopreserve and be careful about
what blocksize you use. Small files, small blocks.

I find system time goes up, user time down, overall time spent scribbling
goes slightly down on average. That really depends on your usage. It
requires tuning to work well.

Also, when I had this deadlock, my one mounted reiser LV (what had almost
no files open) lost MUCH data after a reiserfsck. Practically EVERYTHING
in every dir. Needless to say, mke2fs /dev/vg00/opt...

> P.S.: does somebody know about any Linux JFSes available for testing?

I only know of the LFS projects out there...

-Shawn
<=========== America Held Hostage ===========>
   Day 2218 for the poor and the middle class. 
   Day 2237 for the rich and the dead.
   704 days remaining in the Raw Deal.
<============================================> 

Re: [linux-lvm] User count, general question for all y'all

[Heinz Mauelshagen]

> 
> On Tue, 16 Feb 1999, Heinz Mauelshagen wrote:
> 
> > BTW: i have not been able to test any performance boosts with reiserfs;
> >      reiserfsck is far to slow to be usefull
> 
> Reiser is slower unless you use notails,nopreserve and be careful about
> what blocksize you use. Small files, small blocks.
> 
> I find system time goes up, user time down, overall time spent scribbling
> goes slightly down on average. That really depends on your usage. It
> requires tuning to work well.
> 
> Also, when I had this deadlock, my one mounted reiser LV (what had almost
> no files open) lost MUCH data after a reiserfsck. Practically EVERYTHING
> in every dir. Needless to say, mke2fs /dev/vg00/opt...
> 
> > P.S.: does somebody know about any Linux JFSes available for testing?
> 
> I only know of the LFS projects out there...
> 


Thank you for the information,

i thought there could be other JFS activities being pretty far out there...

IMHO it is a question of 9999 small Linux installations vs. the big one
driving JFSes for Linux coming up ;*)

Heinz
--

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Systemmanagement C/S                             Deutsche Telekom AG
                                                 Entwicklungszentrum Darmstadt
Heinz Mauelshagen                                Otto-Roehm-Strasse 71c
Senior Systems Engineer                          Postfach 10 05 41
                                                 64205 Darmstadt
mge@ez-darmstadt.telekom.de                      Germany
                                                 +49 6151 886-425
                                                          FAX-386
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: [linux-lvm] User count, general question for all y'all

[Shawn Leas]

On Tue, 16 Feb 1999, Stephen Costaras wrote:

> I signed up to the list to get a feel as to how stable lvm is and
> what pitfalls I would cross in using it.  The servers here are used
> in a production setting so I am on the reluctant side of putting in
> 'bleeding edge' stuff.
> 
> Steve

When you think about it, the fundamental role of LVM has
little opportunity to introduce instability where it wasn't
there before. (Maybe Heinz can help me here)

It's simply offereing userland a virtual view of disk (or
some non-volatile storage), and in doing so has the job of
translating logical addresses to physical ones, etc etc. Well,
that and locking some things here and there at times... But
then, that's more for LVM meta data, because LVM sits on top
of already existing kernel stuff, proven stable.

I guess locking could be a concern, but I get the feel,
having used LVM, that Heinz has done a production quality job
with everything.

Go ahead and try it out on a more developmenty platform,
I think you'll find it as stable as a non-lvm one.

-Shawn
<=========== America Held Hostage ===========>
   Day 2218 for the poor and the middle class. 
   Day 2237 for the rich and the dead.
   704 days remaining in the Raw Deal.
<============================================> 

Re: [linux-lvm] User count, general question for all y'all

[Heinz Mauelshagen]

> 
> On Tue, 16 Feb 1999, Stephen Costaras wrote:
> 
> > I signed up to the list to get a feel as to how stable lvm is and
> > what pitfalls I would cross in using it.  The servers here are used
> > in a production setting so I am on the reluctant side of putting in
> > 'bleeding edge' stuff.
> > 
> > Steve
> 
> When you think about it, the fundamental role of LVM has
> little opportunity to introduce instability where it wasn't
> there before. (Maybe Heinz can help me here)

Shawn is right with this because the LVM driver does a quite simple
remapping job these days based on tables loaded by user land tools
(vgchange, vgcreate, lvcreate etc.)

I have plans for the future (RAID > 0) which will make the driver more complex
in means of remapping. But this will only be if you choose to create
RAID > 0 logical volumes then and will have no impact on linear or striped
logical volumes possible today.

BTW: i don't know what future really means 8*)))

> 
> It's simply offering userland a virtual view of disk (or
> some non-volatile storage), and in doing so has the job of
> translating logical addresses to physical ones, etc etc.

See my statement above.

> Well,
> that and locking some things here and there at times... But
> then, that's more for LVM meta data, because LVM sits on top
> of already existing kernel stuff, proven stable.

Yes. See lvm_map() in drivers/block/lvm.c for the remapping stuff.

The only problem being forced to the surface by LVM caused
by the storage amount it can deal with (up to 1 Terabyte today) is mass
i/o instability in some Linux kernel versions.
I watched that problem with several 2.1.x and logical volumes of 20-30 GB.

This is basically fixed with 2.2.1 today.

But 2.2.1 suffers from being to agressive in stealing memory pages
for the buffer cache which leads for eg. to a mke2fs of a big logical volume
(or a big partition) forcing out process pages 8*(

An artivicial limit for buffer memory or more agressive bdflush() activities
hacked into 2.2.1 (fs/buffer.c) helps with this but is not the right
solution for the medium term.

Stephen Tweedie and others are working on this and i think the problem
will disapear in 2.2.2.

> 
> I guess locking could be a concern, but I get the feel,
> having used LVM, that Heinz has done a production quality job
> with everything.

Thanks and thank you all for enhancing its quality!

> 
> Go ahead and try it out on a more developmenty platform,
> I think you'll find it as stable as a non-lvm one.
> 

Agreed ;*)

Heinz

--

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Systemmanagement C/S                             Deutsche Telekom AG
                                                 Entwicklungszentrum Darmstadt
Heinz Mauelshagen                                Otto-Roehm-Strasse 71c
Senior Systems Engineer                          Postfach 10 05 41
                                                 64205 Darmstadt
mge@ez-darmstadt.telekom.de                      Germany
                                                 +49 6151 886-425
                                                          FAX-386
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: [linux-lvm] User count, general question for all y'all

[Heinz Mauelshagen]

> 
> Are we all (on this list) actively using lvm, and in what
> configuration? How many of us are there?
> 

I use it 8*)))

> Also, has anyone encountered deadlocks during heavy IO?

Yes, i did from the early 2.1.x times on.
The question is which Linux version you are using _and_
which aic7xxx version if any.

It turned out to be a mass i/o lock which can be seen without LVM too.
Things turned out to be better in the later 2.1.x and in 2.2.x.

A problem still stays which Stephen Tweedie and others are addressing right now:
with large quantity i/o 2.2.1 preferes allocation of buffer pages vs.
keeping process pages in memory making the system _very_ lazy ;*)


> I'm
> only using LVM at this point as a method of easy partitioning.
> No mirroring, etc. I suspect reiserfs is to blaim on my system,
> but I just thought I'd ask.
> 

Mybe it's a big mkreiserfs forcing the metioned behaviour?

Heinz
--

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Systemmanagement C/S                             Deutsche Telekom AG
                                                 Entwicklungszentrum Darmstadt
Heinz Mauelshagen                                Otto-Roehm-Strasse 71c
Senior Systems Engineer                          Postfach 10 05 41
                                                 64205 Darmstadt
mge@ez-darmstadt.telekom.de                      Germany
                                                 +49 6151 886-425
                                                          FAX-386
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-


--

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Systemmanagement C/S                             Deutsche Telekom AG
                                                 Entwicklungszentrum Darmstadt
Heinz Mauelshagen                                Otto-Roehm-Strasse 71c
Senior Systems Engineer                          Postfach 10 05 41
                                                 64205 Darmstadt
mge@ez-darmstadt.telekom.de                      Germany
                                                 +49 6151 886-425
                                                          FAX-386
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

[linux-lvm] User count, general question for all y'all

[Shawn Leas]

Are we all (on this list) actively using lvm, and in what
configuration? How many of us are there?

Also, has anyone encountered deadlocks during heavy IO? I'm
only using LVM at this point as a method of easy partitioning.
No mirroring, etc. I suspect reiserfs is to blaim on my system,
but I just thought I'd ask.

LVM is soooo kewl!

-Shawn
<=========== America Held Hostage ===========>
   Day 2217 for the poor and the middle class. 
   Day 2236 for the rich and the dead.
   705 days remaining in the Raw Deal.
<============================================>