Anonymous_User Absent Member.
Absent Member.
1384 views

Is expanding pool degrade the IO performance


Hi All,

Question regarding to expanding the pool, will you suffer IO performance
if you are going to expand the existing pool with more disks? Expanding
the pool is not the same as striping, correct?

Thanks,

Steve



Labels (2)
0 Likes
13 Replies
Knowledge Partner
Knowledge Partner

Re: Is expanding pool degrade the IO performance

Hi,

Steve wrote:
>
> Hi All,
>
> Question regarding to expanding the pool, will you suffer IO performance
> if you are going to expand the existing pool with more disks?


It is possibly, yes. The amount seems to vary a lot, from non-noticeable
to hefty. *If possible* I would try to avoid it. E.G, some people tend
to expand their pools a lot, in small increments. I wouldn't do that. I
only span pools when all alternatives aren't feasible.

> Expanding
> the pool is not the same as striping, correct?


Well, some would still call it striping, just the sripe sets are much
bigger than one usually thinks of when "striping" is used. But of
course, the pool data *is* striped over both segments, and losing one
segment means losing the whole pool.

CU,
--
Massimo Rosen
Novell Product Support Forum Sysop
No emails please!
http://www.cfc-it.de
CU,
--
Massimo Rosen
Micro Focus Knowledge Partner
No emails please!
http://www.cfc-it.de
0 Likes
Anonymous_User Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance



Is expanding more like concatenated rather stripping? Is the same apply to
Solaris or Linux volume manager or Windows dynamic disk when the volume is
expand by adding a new disk to the existing volume?


Thanks,

Steve

0 Likes
Knowledge Partner
Knowledge Partner

Re: Is expanding pool degrade the IO performance

Hi,

Steve wrote:
>
> Is expanding more like concatenated rather stripping?


Is there *really* a difference? But no, I wouldn't call in concatenated.
As soon as multiple segments exist, both will be used, even if there's
still free space on one.

CU,
--
Massimo Rosen
Novell Product Support Forum Sysop
No emails please!
http://www.cfc-it.de
CU,
--
Massimo Rosen
Micro Focus Knowledge Partner
No emails please!
http://www.cfc-it.de
0 Likes
Anonymous_User Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance



So, what is the best way to increase the disk space if the storage group
does not want to create a large LUN? I am looking at at least 2TB of
volume in size. Currently, we are at 1TB and we are closed to being full
and we wanted to increase the size of the existing volume.


Thanks,

Steve



0 Likes
Knowledge Partner
Knowledge Partner

Re: Is expanding pool degrade the IO performance

Hi,

Steve wrote:
>
> So, what is the best way to increase the disk space if the storage group
> does not want to create a large LUN? I am looking at at least 2TB of
> volume in size. Currently, we are at 1TB and we are closed to being full
> and we wanted to increase the size of the existing volume.


1TB is a damn lot for one volume. I would look at multiple volumes
really, and/or maybe DFS. But if necessary, go ahead and span it. But
don't slice it too small, and when you see an opportunity not to....

CU,
--
Massimo Rosen
Novell Product Support Forum Sysop
No emails please!
http://www.cfc-it.de
CU,
--
Massimo Rosen
Micro Focus Knowledge Partner
No emails please!
http://www.cfc-it.de
0 Likes
Anonymous_User Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

Massimo Rosen wrote:


> 1TB is a damn lot for one volume.



IIRC with the launch of NW5.0 there was a Novell Article telling about the
advantages of NSS, and in this article NSS was specified to support volumes up to
8EB? ( = 8000 x 1TB)

Well, I searched documentation and failed to find this number (8EB), just an
article from "outside of Novell":

http://articles.techrepublic.com.com/5100-6345-1048521.html


So you recommend *NOT* to use volumes larger than 1TB? What about max.
recommended pool size?


Any differences between netware / linux-OES?


Regards, Rudi.

--
IT-Beratung Rudolf Thilo
Schweinfurter Str. 131
97464 Niederwerrn
t: +49 (0)9721/6464840
f: +49 (0)9721/6464841
m: +49(0)171/685 9 685
0 Likes
Marcel_Cox Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

Rudolf Thilo wrote:

>IIRC with the launch of NW5.0 there was a Novell Article telling about the
>advantages of NSS, and in this article NSS was specified to support
>volumes up to
>8EB? ( = 8000 x 1TB)
>
>Well, I searched documentation and failed to find this number (8EB), just
>an
>article from "outside of Novell":


Unfortunately, some poor design choices including the limiting thelseves
to the old style DOS partitioning sceheme has leead to the fact that NSS
is severely limited in its overall capacity. IT can only handle disks up
to 2TB and an overall pool size up to 8TB (when spanning 4 disks of 2TB).
This is far from the 8EB that could have been achieved with a more
openminded full 64 bit approach at all levels.

--
Marcel Cox
Discover the new Novell forums at http://forums.novell.com
0 Likes
Anonymous_User Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

Marcel Cox wrote:

> Unfortunately, some poor design choices including the limiting thelseves to the
> old style DOS partitioning sceheme has leead to the fact that NSS is severely
> limited in its overall capacity. IT can only handle disks up to 2TB and an
> overall pool size up to 8TB (when spanning 4 disks of 2TB). This is far from
> the 8EB that could have been achieved with a more openminded full 64 bit
> approach at all levels.


Too sad to hear that. Do I remember correctly that when introducing NSS the specs
given to the people were telling a Pool/Volume size up to 8EB?


Well, as disks sizes grow and grow it's just a question of time, that NSS will
not be able to handle the full size of *ONE* disk (2TB limit)

So let me ask if you've heard anything about some plans to get rid of these
compatibility "chains" ?


Regards, Rudi.

--
IT-Beratung Rudolf Thilo
Schweinfurter Str. 131
97464 Niederwerrn
t: +49 (0)9721/6464840
f: +49 (0)9721/6464841
m: +49(0)171/685 9 685
0 Likes
Marcel_Cox Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

Increasing the NSS size limits is a much asked for enhancement and is
something Novell is seriously looking into how to do it. However if the
limits will ever change, it will surely only be on OES/Linux and not on
NetWare.

--
Marcel Cox (Sysop)
Discover the new Novell forums at http://forums.novell.com
0 Likes
Anonymous_User Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

Marcel Cox wrote:

> Increasing the NSS size limits is a much asked for enhancement and is something
> Novell is seriously looking into how to do it. However if the limits will ever
> change, it will surely only be on OES/Linux and not on NetWare.


Surely?

EOL for NetWare 6.5:

General Support: 07 Mar 2010
Extended Support: 07 Mar 2012
Self Support: 07 Mar 2015


Even for the end of General Support we might see Laptop(!) disks with over 2TB.

I cannot imagine, that Novell will call this "Support" if NW6.5.x cannot even
handle a single disk's size!!!

On the other hand: It's Novell <scnr>

So they might sit and wait until this ("cannot handle a single disk's size") *IS*
fact before they will start to act in hectic manner.



This idea is too sad 😞 so say it's not true. Please!


Regards, Rudi.

--
IT-Beratung Rudolf Thilo
Schweinfurter Str. 131
97464 Niederwerrn
t: +49 (0)9721/6464840
f: +49 (0)9721/6464841
m: +49(0)171/685 9 685
0 Likes
Marcel_Cox Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

Rudolf Thilo wrote:

>Surely?


Yes.

Support means problem support but not development of new features.
Support means that people can keep there existing systems running, but
doesn't mean they should be encouraged to use NetWare for new installations.


--
Marcel Cox (Sysop)
Discover the new Novell forums at http://forums.novell.com
0 Likes
Anonymous_User Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

Marcel Cox wrote:

> Rudolf Thilo wrote:
>
> > Surely?

>
> Yes.
>
> Support means problem support but not development of new features.


NW5.0 was 1999, was it?

So started almost 10 years ago with "up to 8 Exa" and ending with, ...

well, OK, forget it: I got it... 😞


Rudi.


--
IT-Beratung Rudolf Thilo
Schweinfurter Str. 131
97464 Niederwerrn
t: +49 (0)9721/6464840
f: +49 (0)9721/6464841
m: +49(0)171/685 9 685
0 Likes
ecyoung Absent Member.
Absent Member.

Re: Is expanding pool degrade the IO performance

You could still submit an enhancement request considering that Novell
defines "General Support" as having "The ability to request product and
feature enhancements". Somehow I don't have much faith that it would even
be looked at.

-Eric

"Rudolf Thilo" <spamfrei@nurfuerspam.de> wrote in message
news:L3wjj.20842$k06.18568@kovat.provo.novell.com...
> Marcel Cox wrote:
>
>> Rudolf Thilo wrote:
>>
>> > Surely?

>>
>> Yes.
>>
>> Support means problem support but not development of new features.

>
> NW5.0 was 1999, was it?
>
> So started almost 10 years ago with "up to 8 Exa" and ending with, ...
>
> well, OK, forget it: I got it... 😞
>
>
> Rudi.
>
>
> --
> IT-Beratung Rudolf Thilo
> Schweinfurter Str. 131
> 97464 Niederwerrn
> t: +49 (0)9721/6464840
> f: +49 (0)9721/6464841
> m: +49(0)171/685 9 685



0 Likes
The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.