Wouldn''t it be great if the opensolaris community creates a fs fzfs (for flash zfs) that could be the the filesystem for SDXC cards? Two main point for this are already there Sun is a member of SD Card Association and the code writting smart on a flash should be there as well with stuff of l2arc. I think that could give the opensolaris a more visiblie/ markting in the IT space and would bring zfs support to even more OSes. Please correct me if I am wrong with anything? Thanks for reading, f. -- This message posted from opensolaris.org
On Sun, Jan 11 at 5:00, Fabian W?rner wrote:>Wouldn''t it be great if the opensolaris community creates a fs fzfs (for flash zfs) that >could be the the filesystem for SDXC cards? Two main point for this are already there >Sun is a member of SD Card Association and the code writting smart on a flash should be there as well with stuff of l2arc. >I think that could give the opensolaris a more visiblie/ markting in the IT space and >would bring zfs support to even more OSes. >Please correct me if I am wrong with anything? >Thanks for reading, >f.My impression is not that other OS''s aren''t interested in ZFS, they are, it''s that the licensing restrictions limit native support to Solaris, BSD, and OS-X. If you wanted native support in Windows or Linux, it would require a significant effort from Sun. -- Eric D. Mudama edmudama at mail.bounceswoosh.org
On Sun, 11 Jan 2009, Eric D. Mudama wrote:> > My impression is not that other OS''s aren''t interested in ZFS, they > are, it''s that the licensing restrictions limit native support to > Solaris, BSD, and OS-X.Perhaps the philosophical issues of the "other OS''s" (i.e. Linux) are more significant than the actual licensing issues. Many/most Linux users could legally use a native optimized kernel implementation of Sun ZFS if it was offered to them to do so. GPLv2 only adds restrictions when copying binaries. A pure source based distribution like Gentoo has hardly any issues at all. Bob =====================================Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
Bob Friesenhahn wrote:> On Sun, 11 Jan 2009, Eric D. Mudama wrote: >> My impression is not that other OS''s aren''t interested in ZFS, they >> are, it''s that the licensing restrictions limit native support to >> Solaris, BSD, and OS-X. > > Perhaps the philosophical issues of the "other OS''s" (i.e. Linux) are > more significant than the actual licensing issues. Many/most Linux > users could legally use a native optimized kernel implementation of > Sun ZFS if it was offered to them to do so. GPLv2 only adds > restrictions when copying binaries. A pure source based distribution > like Gentoo has hardly any issues at all.Nobody in their right mind is using Gentoo. If you want it in Linux then it has to be a proper GPL compliant effort. I for one would like this to happen. Tom
On Sun, 11 Jan 2009 20:28:36 +0000 Tom Bird <tom at marmot.org.uk> wrote:> Bob Friesenhahn wrote: > > On Sun, 11 Jan 2009, Eric D. Mudama wrote: > >> My impression is not that other OS''s aren''t interested in ZFS, they > >> are, it''s that the licensing restrictions limit native support to > >> Solaris, BSD, and OS-X. > > > > Perhaps the philosophical issues of the "other OS''s" (i.e. Linux) > > are more significant than the actual licensing issues. Many/most > > Linux users could legally use a native optimized kernel > > implementation of Sun ZFS if it was offered to them to do so. > > GPLv2 only adds restrictions when copying binaries. A pure source > > based distribution like Gentoo has hardly any issues at all. > > Nobody in their right mind is using Gentoo.Based on ...?> If you want it in Linux then it has to be a proper GPL compliant > effort.Why should we want it in linux? It may be that one day linux wants it in linux ;-)> I for one would like this to happen.GPL sucks. At least, that''s my opinion. GPL dominion sucks even bigger. -- Dick Hoogendijk -- PGP/GnuPG key: 01D2433D + http://nagual.nl/ | SunOS sxce snv104 ++ + All that''s really worth doing is what we do for others (Lewis Carrol)
yeah, this is ZFS turf. what Linux folks really care may not be what z-open folks really care... open folks are not just all in one camp. I like you open folks, much much more than the L-open folks. best, z ----- Original Message ----- From: "dick hoogendijk" <dick at nagual.nl> To: <zfs-discuss at opensolaris.org> Sent: Sunday, January 11, 2009 3:47 PM Subject: Re: [zfs-discuss] SDXC and the future of ZFS> On Sun, 11 Jan 2009 20:28:36 +0000 > Tom Bird <tom at marmot.org.uk> wrote: > >> Bob Friesenhahn wrote: >> > On Sun, 11 Jan 2009, Eric D. Mudama wrote: >> >> My impression is not that other OS''s aren''t interested in ZFS, they >> >> are, it''s that the licensing restrictions limit native support to >> >> Solaris, BSD, and OS-X. >> > >> > Perhaps the philosophical issues of the "other OS''s" (i.e. Linux) >> > are more significant than the actual licensing issues. Many/most >> > Linux users could legally use a native optimized kernel >> > implementation of Sun ZFS if it was offered to them to do so. >> > GPLv2 only adds restrictions when copying binaries. A pure source >> > based distribution like Gentoo has hardly any issues at all. >> >> Nobody in their right mind is using Gentoo. > > Based on ...? > >> If you want it in Linux then it has to be a proper GPL compliant >> effort. > > Why should we want it in linux? It may be that one day linux wants it > in linux ;-) > >> I for one would like this to happen. > > GPL sucks. At least, that''s my opinion. GPL dominion sucks even bigger. > > -- > Dick Hoogendijk -- PGP/GnuPG key: 01D2433D > + http://nagual.nl/ | SunOS sxce snv104 ++ > + All that''s really worth doing is what we do for others (Lewis Carrol) > _______________________________________________ > zfs-discuss mailing list > zfs-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
On 11-Jan-09, at 3:28 PM, Tom Bird wrote:> Bob Friesenhahn wrote: >> On Sun, 11 Jan 2009, Eric D. Mudama wrote: >>> My impression is not that other OS''s aren''t interested in ZFS, they >>> are, it''s that the licensing restrictions limit native support to >>> Solaris, BSD, and OS-X. >> >> Perhaps the philosophical issues of the "other OS''s" (i.e. Linux) are >> more significant than the actual licensing issues. Many/most Linux >> users could legally use a native optimized kernel implementation of >> Sun ZFS if it was offered to them to do so. GPLv2 only adds >> restrictions when copying binaries. A pure source based distribution >> like Gentoo has hardly any issues at all. > > Nobody in their right mind is using Gentoo. >Hmmm... a lot of seasoned sysadmins would disagree. A source based distribution has some significant advantages; I would not be surprised if that concept outlives binary packaging and its attendant dependency hell. --Toby> If you want it in Linux then it has to be a proper GPL compliant > effort. > > I for one would like this to happen. > > Tom > _______________________________________________ > zfs-discuss mailing list > zfs-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
On Sun, 11 Jan 2009, Toby Thain wrote:>> >> Nobody in their right mind is using Gentoo. > > Hmmm... a lot of seasoned sysadmins would disagree. A source based > distribution has some significant advantages; I would not be surprised if > that concept outlives binary packaging and its attendant dependency hell.I had Gentoo here and have lived to tell the tale. It would be a tall tale indeed if I was to claim to have recieved some sort of scars or undue stress from the experience, or even significant discomfort. The main issue was with burning CPU building new versions of software, but that is much less of an issue now. I do not claim to be in right mind without someone paying me money to do so. Regardless, I am not aware of anything preventing someone from creating a distributable source patch to insert ZFS into the Linux kernel, as long as the end user builds and installs their own kernel. While it may be difficult for some Linux users to understand, the source code is actually available for the Linux kernel. Bob =====================================Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
You are receiving this email as you are a subscriber with a standard
membership. To update your membership preferences, please see the links at
the bottom of this message.
January 5 - January 11
Volume 131 Issue 1
Customize My News Report or Change My Profile Login to Web Site
Unsubscribe Change Report Format to Text View the archive By Date
By Section
This report built for:
the z-open folks Compliments of:
System News
production at systemnews.com
Sections in this issue that you have enabled:
News, MySQL, Performance, Networking, BigAdmin, Solaris,
OpenSolaris
Sections available in this issue that you have disabled:
Features, Top10, Events, Security, Service, Software, Servers,
Developer, Sysadmin
Click section name to enable it
The United States Marine Corps has selected the Sun Identity
Management Suite and Sun has purchased Q-layer, a firm whose technology
simplifies cloud management and allows users to quickly provision and deploy
applications. Learn more about Sun and cloud computing from Dave Douglas.
Read more Good News from Sun marketing.
See how better SAP performs on Solaris (with unicode) over Linux
(non-unicode), on very similar hardware and how much Siebel CRM users can
save by upgrading to the T5440.
We all work hard to create data, but what''s your policy when it
come
time to destroy data?
Learn about Solaris and security in the upcoming book "Solaris
Security Essentials". A new release of LDoms, 1.1, is now available. Read
about the security advantages that stem from the fine-grained resource
management capabilities of Solaris Zones.
Turn this section off News
[21156] Sun Identity Management Suite to Protect U.S. Marine Corps
Will Serve as Identity and Access Management Solutions for USMC
Systems
The Sun Identity Management Suite is the solution of choice for the
United States Marine Corps (USMC) to protect over 400,000 Marines in
classified and unclassified networks. The Sun Suite, which includes
Sun''s
Identity Manager, Role Manager, OpenSSO Enterprise, and Directory Server
Enterprise Edition, will provide role-based access to a variety of business
and mission systems to enable data sharing and collaboration in a trusted
environment.
(Get More Information . . .) (Contact Me . . .)
[21163] Cloud Computing Company Q-layer Purchased by Sun
To Join Sun''s New Cloud Computing Business Unit
Sun''s new Cloud Computing Business Unit just added more
technological
options for its customers with the purchase of cloud computing company
Q-layer for an undisclosed amount. Q-layer technology simplifies cloud
management and allows users to quickly provision and deploy applications, a
key component in Sun''s strategy to enable building public and private
clouds.
(Get More Information . . .) (Contact Me . . .)
[21169] Good News from Sun Marketing
January 2009
Each month Sun Marketing publishes the "Good News" about Sun.
This
month there are 19 items and three groups:
a.. Customers Choose Sun: MySQL, Sun Startup Essential program, Sun
Fire Servers, Sun Blade Modular Systems, Sun Secure Global Desktop Software,
Sun GlassFish Enterprise Server, Solaris 10, Sun StorageTek Arrays and
Modular Library Systems
b.. Leading Market Conversations: Open Storage, Cloud Computing,
JavaFX, Disk Storage Systems Sales, Magic Quadrant for Web Access
Management, Sun''s Strategies and Innovation, Asia''s Top Tech
Index,
Financial Performance
c.. Product Reviews Underscore Innovation: MySQL 5.1, OpenSolaris
2008.11, JavaFX, Sun xVM VirtualBox, OpenOffice.org 3.0
A two-page PDF of the "Good News" from Sun is available in
English and
several other languages.
(Get More Information . . .) (Contact Me . . .)
[21153] News Bites - Sun Storage 7000 Family, JavaFX, Solaris Trusted
Extensions, Oracle Analysis, HPC Symposium
Short News Items of Interest to the Sun User Community
a.. Sun Storage 7000 Family Named to Top 10
b.. Top 10 Application Development Products Includes JavaFX
c.. Publications Cover Solaris Trusted Extensions
d.. Techniques on Collecting and Analyzing Oracle Performance Data
e.. Sun Sponsoring HPC Symposium in Ontario
http://blogs.systemnews.com/
(Get More Information . . .) (Contact Me . . .)
[21157] Review of News for 2008 - 1 of 3
50 Most Popular Topics, 900 Most Popular Articles
By accessing the web logs, we have determined that the following list
of about 900 articles were the most popular with subscribers and web site
visitors in 2008.
We have grouped the articles by topic to give you a view of what were
the most important news articles for 2008. That a topic had enough popular
articles to be included in this of topics is in itself significant.
Part 1 of 3
a.. Sun Microsystems, Sun/NetApp Litigation, Jonathan Schwartz,
Servers, Blade Servers, Intel Servers, AMD Servers, SPARC CMT, SPARC64
Servers, Green, Datacenter, Storage, Tape and VTL, ZFS, Open Storage
(Get More Information . . .) (Contact Me . . .)
[21171] Review of News for 2008 - 2 of 3
50 Most Popular Topics, 900 Most Popular Articles
Part 2 of 3
a.. Open Source, HPC, Lustre, Inifniband, Solaris 10, OpenSolaris,
OpenSolaris Project, MySQL, Data Warehouse, Solaris Cluster, Netra, DTrace,
Security, OpenSSO, Java FX
(Get More Information . . .) (Contact Me . . .)
[21172] Review of News for 2008 - 3 of 3
50 Most Popular Topics, 900 Most Popular Articles
Part 3 of 3
a.. Virtualization, Containers, Logical Domains (LDoms), Sun Ray,
Virtual Desktop Infrastructure, Sun xVM, Sun xVM VirtualBox, OpenOffce and
StarOffice, JRuby, Sun Studio, VMware, Windows, Linux, GlassFish, NetBeans,
Mobile, Identity Management, Sun Java System, Java
(Get More Information . . .) (Contact Me . . .)
Turn this section off MySQL
[21083] Mark Matthews on MySQL Query Analyzer
What It Does to Make Life Easier for the DBA
Mark Matthews is one of the architects of the MySQL Query Analyzer
and, therefore, a logical choice as an interview subject on that aspect of
the MySQL solution.
(Get More Information . . .) (Contact Me . . .)
[21060] ZRM for MySQL Offers Subscriptions for Enterprise and Cluster
Support
Backup and Recovery Solution That''s Easy to Use
Busy database administrators (DBA) using MySQL may be interested in
two recent offerings - the Zmanda Recovery Manager (ZRM) for MySQL
Enterprise and ZRM for MySQL Cluster. ZRM for MySQL offers DBAs an
easy-to-use, flexible, and robust backup and recovery solution. ZRM for
MySQL Enterprise is the backup and recovery solution for business-critical
databases, while ZRM for MySQL Cluster is designed for the demands of MySQL
Cluster CGE.
(Get More Information . . .) (Contact Me . . .)
Turn this section off Performance
[21006] Making Sense of IOPS Results
Don''t Forget About Latency
Sun Engineer Brendan Gregg on the Fishworks team offers some useful
clarifications concerning the stated results of IOPS testing in his blog "A
quarter million NFS IOPS." He includes screenshots from Analytics that show
sustained NFS read ops/sec from DRAM. Starting with NFSv3 and then provides
explanations.
(Get More Information . . .) (Contact Me . . .)
[21022] How Does SAP Perform on Solaris x64 Compared to Linux?
Unicode Versus Non-Unicode
The two-tier SAP-SD standard application benchmark provides results
for whether SAP on Solaris x64 performs better or worse than on Linux or
Windows.
Benchmarks are performed by HP using SAP ERP 6.0 (2005) non-Unicode
with Oracle 10g RDBMS and by Sun using SAP ERP 6.0 (2005) Unicode with MaxDB
7.6 RDBMS.
(Get More Information . . .) (Contact Me . . .)
[21081] Boost Your ROI with Sun Optimized Oracle CRM Solution and the
Sun SPARC Enterprise T5440 Server
14,000 Users on a 4 RU Package at a Fourth of the Cost
Paul Milo, group manager, Oracle Sun Market Development, has a
persuasive marketing message for users thinking about upgrading their legacy
CRM solutions and the savings that such a move can deliver. For example, he
points out that there is potentially a 1900% ROI in reduced operating costs
for users who upgrade legacy CRM solutions, such as Siebel 6 or Siebel 7, to
Oracle''s Siebel 8 CRM Platform Sizing and Performance Program (PSPP)
running
on the Sun SPARC Enterprise T5440 Server, which can provide savings of as
much as $5 million over five years and lower energy usage by up to 90%.
(Get More Information . . .) (Contact Me . . .)
Turn this section off Networking
[21063] Sun Storage FC Switch Model 5802
Build Next Generation SANs for Today''s Infrastructure
The Sun Storage Fibre Channel (FC) Switch Model 5802 offers 8Gb FC,
providing more bandwidth for SAN while also extending the life of existing
SAN infrastructure since it is backwards compatible with 4Gb and 2Gb devices
and optics. The stacking technology used for this switch allows it to scale
from 8 to 120 FC device ports and connect up to 6 FC switches into a single
manageable switch stack.
(Get More Information . . .) (Contact Me . . .)
[21064] Brocade Data Center Fabric Manager (DCFM)
Manage Brocade Products with Less Cost, Complexity
Sun has announced the Brocade Data Center Fabric Manager (DCFM)
software, which simplifies infrastructure management with the performance
and scalability required for the majority of the world''s data centers.
With
DCFM, all Brocade products, whether based on the Fabric Operating System
(FOS) or the M-EOS operating system (McData EFCM), can be managed with less
cost and complexity. DCFM is available as two different products: DCFM
Professional and DCFM Enterprise.
(Get More Information . . .) (Contact Me . . .)
Turn this section off BigAdmin
[21179] What''s New on BigAdmin
As of January 8th, 2008
Articles which have been posted to BigAdmin in the last week include:
a.. System Administration Resources in Japanese and Other Languages
b.. Community Article: Upgrading to the Solaris 10 10/08 OS for
SPARC Platforms
c.. Community Script: ut_gather 1.4.6
d.. Community Script: grepr (recursive grep)
http://www.sun.com/bigadmin
(Get More Information . . .) (Contact Me . . .)
Turn this section off Solaris
[21072] An Accelerated Introduction to Solaris 10: Part 1
By Ben Rockwood
This "whirl wind tour" of Solaris OS fills in the gaps from the
ground up.
a.. Solaris File system Layout
b.. The Automounter
c.. Sudo and Root
d.. NFS Shares and /etc/exports
e.. Basic Monitoring
f.. Software and Packages
g.. Managing Services
h.. Managing Disks
(Get More Information . . .) (Contact Me . . .)
[21160] Stricter Solaris Patch Entitlement Implementation
Support Contract Required for Access to Many Patches
The Sun Weblog "Patch Corner" reports a stricter Solaris patch
entitlement implementation will be going into effect to ensure it matches
the Solaris patch entitlement policy. Under the new entitlement
implementation, 18% of Solaris OS patches will remain available without a
support contract, while the remainder will require a valid support contract
in order to access them. The new implementation will roll out in phases,
starting the second full week of January.
(Get More Information . . .) (Contact Me . . .)
Turn this section off OpenSolaris
[21026] A How-to on Using OpenSolaris as a File Server
ZFS File System, Relatively Fast Processor Make It Happen
Dr. Oliver Diedrich, writing for Heise Open Source Forum, discusses
the subject of using OpenSolaris as a file server. Both a very fast SMB/CIFS
server and the modern functionality of the ZFS file system make OpenSolaris
a good choice when setting up a LAN file server, he contends. For a
relatively modest cost, Diedrich asserts, one can build a storage system
that promises more data safety and especially a much higher performance than
a simple network hard disk using OpenSolaris.
(Get More Information . . .) (Contact Me . . .)
[21043] OpenSolaris 2008.11 Installation FAQ
By Roman Strobl
This FAQ defines common questions that users may have for the
installation of OpenSolaris.
The how-to topics covered include virtual box, driver device utility,
installation guide, VMWare, disk, partitions, bare metal, boot manager,
Linux, Mac installations, installation blogs, user guides, and
documentation.
(Get More Information . . .) (Contact Me . . .)
[21066] Sun News Radio Covers OpenSolaris 2008.11
Latest Updates on its New Data Center and Storage Components
Charlie Boyle, Director of Solaris Marketing, Sun Microsystems
provides the latest updates on the new datacenter and storage components of
OpenSolaris 2008.11. Boyle discusses the innovative feature, Time Slider;
the new repository packages introduced; developments in desktop experience;
wireless networking support; and suspend/resume capabilities.
(Get More Information . . .) (Contact Me . . .)
About System News, Inc. (SNI)
SNI, a privately held company since its inception in 1998, provides a
custom news service to people who use Sun Microsystems Products.
SNI makes private label editions of its custom news report service for
sales teams who work for Sun or an authorized Sun Microsystems Resellers
(Partners Advantage Program).
SNI also offer ''go-to-market'' programs for Partners
(ISVs, IHVs, etc.)
who want to reach some or all of our 45,000 subscribers.
System News, Inc., as an ASP, is a member of Sun''s Partner
Advantage
Program.
Contact:
(954) 973-9000 info at systemnews.com More about SNI
http://sun.systemnews.com
Hits : 10 20 50 100 200 Age : none a year 6 months 90 days 30 days
2 weeks 1 week
You last updated your settings on: January 8th, 2009.
How to customize this report:
a.. Enable/Disable sections based on your interests
b.. Choose your preferred news report format
c.. Change your contact information, including email address
d.. Suspend or resume this service
e.. Discontinue this service by deleting your account
To do so, please update your settings.
Additional capabilities will be added in the future.
You are encouraged to send this invite URL to your
colleagues/customers so they can also take advantage of this service:
http://sun.systemnews.com/subscribe/invite
Do NOT forward this email, as it contains your account information.
You may not re-distribute, re-use, re-purpose or translate the
contents of this news report, or the corresponding Web site, in part or in
whole, without license or explicit permission.
This news report service is produced by System News, Inc.
+1 (954) 973 9000 info at systemnews.com
934 N. University Drive, #305
Coral Springs, FL 33071 USA
See http://www.systemnews.com for details. Sun, Sun Microsystems, the
Sun logo, CoolThreads, Java, JavaFX, JVM, NetBeans, OpenSolaris, Solaris,
Sun Blade, Sun Fire, Sun Ray, StarOffice, Sun StorEdge, Sun StorageTek,
iForce and iPlanet are trademarks or registered trademarks of Sun
Microsystems, Inc. in the United States and other countries.
All SPARC trademarks are used under license and are trademarks or
registered trademarks of SPARC International, Inc. in the United States and
other countries. Products bearing SPARC trademarks are based upon an
architecture developed by Sun Microsystems, Inc.
UNIX is a registered trademark in the United States and other
countries, exclusively licensed through The Open Group.
Mozilla is a trademark or registered trademark of Netscape
Communications Corporation.
OpenGL is a registered trademark of Silicon Graphics, Inc. (SGI).
(c) 2008 System News, Inc. All rights reserved.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 15868 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 868 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0001.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 2826 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0002.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 993 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0003.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 807 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0004.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 9182 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 5402 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0001.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 8065 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0002.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 5377 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0003.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 4624 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0004.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 4383 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0005.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 3652 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0006.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 6022 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0007.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 19936 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 3499 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0008.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 4288 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0009.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 2680 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0010.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 4787 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090111/44734eed/attachment-0011.jpe>
>My impression is not that other OS''s aren''t interested in ZFS, they >are, it''s that the licensing restrictions limit native support to >Solaris, BSD, and OS-X. > >If you wanted native support in Windows or Linux, it would require a >significant effort from Sun.Why is that a problem for Windows? Linux, yes, but if they want they can change that. Casper
"Eric D. Mudama" <edmudama at bounceswoosh.org> wrote:> My impression is not that other OS''s aren''t interested in ZFS, they > are, it''s that the licensing restrictions limit native support to > Solaris, BSD, and OS-X.The BDF folks had some cons against the CDDL but after I had a lobger discussion with them, they understood that the CDDL does not hurt them as long as the CDDL is not used for essential siftware core software that cannot be replaced.> If you wanted native support in Windows or Linux, it would require a > significant effort from Sun.The problem is not the CDDL but the way ZFS and the CDDL is discussed in the Linux camp. Even a GPLd ZFS would not change things for Linux.... A GPLd ZFS would however disallow to use it on *BSD and Mac OS X. J?rg -- EMail:joerg at schily.isdn.cs.tu-berlin.de (home) J?rg Schilling D-13353 Berlin js at cs.tu-berlin.de (uni) joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
my post was not to start a discuss gpl<>cddl. It just an idea to promote ZFS and OPENSOLARIS!!!! If it was against anything than against exfat, nothing else!!! -- This message posted from opensolaris.org
Fabian W?rner <fabian.woerner at googlemail.com> wrote:> my post was not to start a discuss gpl<>cddl. > It just an idea to promote ZFS and OPENSOLARIS!!!! > If it was against anything than against exfat, nothing else!!!If you like to promoote ZFS, you need to understand why the party you like to promote it to does not already use it ;-) J?rg -- EMail:joerg at schily.isdn.cs.tu-berlin.de (home) J?rg Schilling D-13353 Berlin js at cs.tu-berlin.de (uni) joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
Joerg Schilling wrote:> Fabian W?rner <fabian.woerner at googlemail.com> wrote: > >> my post was not to start a discuss gpl<>cddl. >> It just an idea to promote ZFS and OPENSOLARIS!!!! >> If it was against anything than against exfat, nothing else!!! > > If you like to promoote ZFS, you need to understand why the party you like > to promote it to does not already use it ;-)And for SDXC, ZFS will probably never be the filesystem of choice. Removable media of this type is mostly used in portable electronic devices, such as cameras, cellphones, etc. All of which are power, CPU, and memory limited. ZFS, while a marvelous filesystem, is incredibly RAM hungry. I suspect it''s CPU profile is also non-trivial for a restricted performance device. Now it _might_ be possible for some of these characteristics to be changed with a code re-write targeting small devices, and probably a feature-limited zpool/zfs version number, but the effort would be non-trivial. -- Carson
>>>>> "tt" == Toby Thain <toby at telegraphics.com.au> writes: >>>>> "j" == JZ <jz at excelsioritsolutions.com> writes:> Nobody in their right mind is using Gentoo. tt> a lot of seasoned sysadmins would disagree. Gentoo makes sense for embedded projects. OpenWRT is arguably source-based too. but Gentoo is extremely clumsy to my view: I find USE flags to be a terrible mistake because they hide bugs and make dependencies no longer automatic. It would be a lot more workable if I had bootable ZFS snapshots in Gentoo to give rollback protection when doing security-fix updates. Also I think they don''t do enough (any?) branching for stable systems---they sort of think they can have ``stable'''' and ``experimental'''' branches both continually updated, with no tagging, all the time, which makes all updates equally risky instead of batching things into major-infrequent and minor-frequent updates. In that sense I guess it''s just like opensolaris. but yeah, that OTness aside, Sun''s deliberately crafting their brand new CDDL license to be incompatible with the GPL isn''t exactly in the spirit of free software. BSD is also not in the GPL camp, but the mainstream of BSD has altered their licenses where possible to add GPL compatibility. The GPL camp moves in the same direction: the GPLv3 added changes to slightly improve license compatibility. That said I don''t really understand why ZFS can''t be a Linux kernel module, since Linus''s ``interpretation'''' of the GPL has almost reduced it to LGPL within the kernel. Why his ``interpretation'''' should have such weight when he doesn''t hold all the copyrights I''ve never understood either, but so far all these big companies distributing binary modules seem to take it as law. This thread seems to say the same: http://groups.google.com/group/zfs-fuse/browse_thread/thread/1219db6af605f792 so, maybe it is not really a license permissiveness issue, so much as a license preference issue that you cannot convince very many talented Linux developers to do free work for you unless you give their efforts the protection of the GPL. The one willing to work without GPL protection only offers enough free time to do the easier FUSE port. :) j> I like you open folks, much much more than the L-open folks. (1) don''t confuse the people with the license. In general I get along better with BSD users, but not their license---I prefer GPL, as user and developer. (2) lots of the people here are Sol10 users, of the stable release. those people are not ``open folks'''' at all. (3) http://www.openbsd.org/papers/opencon06-drivers/mgp00024.html it''s all about the details. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 304 bytes Desc: not available URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090112/04c11261/attachment.bin>
>>>>> "js" == Joerg Schilling <Joerg.Schilling at fokus.fraunhofer.de> writes:js> A GPLd ZFS would however disallow to use it on *BSD and Mac OS js> X. and also Solaris. which is why ZFS would not be GPL''d that way. It''d be a choice of license. I think someone floated the same either/or license as a ``we''ll see'''' possibility for the Java source release. What would happen next is that any improvements Linux developers made might be GPL-only, because they''d have the right to do that. In the case of the atheros code they share with BSD, they supposedly agreed to dual-either/or-license their improvements, but they don''t have to, and they can change their minds. If they didn''t, and Linux got a lively ZFS community, then there''d either be huge duplication of effort or ZFS in Solaris would fall behind to the point that Linux would become the definitive release. I agree that''s far from ideal. But Linux developers have the absolute right to decide how they want to spend their time, and I don''t agree they''re being irrational by preferring to work on btrfs than under the CDDL. I do agree that their _choice_ isn''t a ``problem,'''' but I wouldn''t stonewall while holding my breath waiting for them to suddenly change their minds, either. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 304 bytes Desc: not available URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090112/0d8793e8/attachment.bin>
Carson Gaspar wrote:> Joerg Schilling wrote: > >> Fabian W?rner <fabian.woerner at googlemail.com> wrote: >> >> >>> my post was not to start a discuss gpl<>cddl. >>> It just an idea to promote ZFS and OPENSOLARIS!!!! >>> If it was against anything than against exfat, nothing else!!! >>> >> If you like to promoote ZFS, you need to understand why the party you like >> to promote it to does not already use it ;-) >> > > And for SDXC, ZFS will probably never be the filesystem of choice. > Removable media of this type is mostly used in portable electronic > devices, such as cameras, cellphones, etc. All of which are power, CPU, > and memory limited. ZFS, while a marvelous filesystem, is incredibly RAM > hungry. I suspect it''s CPU profile is also non-trivial for a restricted > performance device. >I''m not sure this is a huge hurdle. There already is a reduced-functionality ZFS implementation in grub, which is quite small. Remember, most of the extreme performance features of ZFS would not necessarily be needed in an embedded system. -- richard
Richard Elling wrote:> Carson Gaspar wrote: > >> Joerg Schilling wrote: >> >> >>> Fabian W?rner <fabian.woerner at googlemail.com> wrote: >>> >>> >>> >>>> my post was not to start a discuss gpl<>cddl. >>>> It just an idea to promote ZFS and OPENSOLARIS!!!! >>>> If it was against anything than against exfat, nothing else!!! >>>> >>>> >>> If you like to promoote ZFS, you need to understand why the party you like >>> to promote it to does not already use it ;-) >>> >>> >> And for SDXC, ZFS will probably never be the filesystem of choice. >> Removable media of this type is mostly used in portable electronic >> devices, such as cameras, cellphones, etc. All of which are power, CPU, >> and memory limited. ZFS, while a marvelous filesystem, is incredibly RAM >> hungry. I suspect it''s CPU profile is also non-trivial for a restricted >> performance device. >> >> > > I''m not sure this is a huge hurdle. There already is a > reduced-functionality > ZFS implementation in grub, which is quite small. Remember, most of the > extreme performance features of ZFS would not necessarily be needed in an > embedded system. >Well, it would be really nice if an open filesystem was adopted on such devices. I''m guessing, but I would suspect exFAT (a.k.a. FAT64) which is the only current contender, isn''t open. -- Andrew
On Mon, 12 Jan 2009, Miles Nordin wrote:> > but yeah, that OTness aside, Sun''s deliberately crafting their brand > new CDDL license to be incompatible with the GPL isn''t exactly in the > spirit of free software. BSD is also not in the GPL camp, but the > mainstream of BSD has altered their licenses where possible to add GPL > compatibility. The GPL camp moves in the same direction: the GPLv3 > added changes to slightly improve license compatibility.After all these years, I am still not sure what is meant by free software. For 99.999% of humanity, "free" means that they don''t have to pay for it. For GPL "free" does not pertain to its use by humans at all. Instead "free" for GPL is about preventing the "enslavement" of the source code itself ("enslavement" means distribution of binaries without source code), as if the source code was a living breathing creature. Due to the ambiguity now associated with the "free", I will prefer the term "open source" and use some other term besides "free" to describe any encumberances caused by the license. Bob =====================================Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/
Does creating ZFS pools on multiple partitions on the same physical drive still run into the performance and other issues that putting pools in slices does?
Le 12 janv. 09 ? 17:39, Carson Gaspar a ?crit :> Joerg Schilling wrote: >> Fabian W?rner <fabian.woerner at googlemail.com> wrote: >> >>> my post was not to start a discuss gpl<>cddl. >>> It just an idea to promote ZFS and OPENSOLARIS!!!! >>> If it was against anything than against exfat, nothing else!!! >> >> If you like to promoote ZFS, you need to understand why the party >> you like >> to promote it to does not already use it ;-) > > And for SDXC, ZFS will probably never be the filesystem of choice. > Removable media of this type is mostly used in portable electronic > devices, such as cameras, cellphones, etc. All of which are power, > CPU, > and memory limited. ZFS, while a marvelous filesystem, is incredibly > RAM > hungry. I suspect it''s CPU profile is also non-trivial for a > restricted > performance device.I have not looked at it recently but for any access greater than ~ 16K ZFS was more efficient than UFS. It''s just one partial data point but the conventional wisdom that ZFS will use more cpu is not an absolute truth. Even more so for RAM, ZFS with 128K record make efficient use of metadata. The only ram it needs to operation is 10 seconds of of your workload''s throughput and that can be tuned down in appliances. -r> Now it _might_ be possible for some of these characteristics to be > changed with a code re-write targeting small devices, and probably a > feature-limited zpool/zfs version number, but the effort would be > non-trivial. > > -- > Carson > _______________________________________________ > zfs-discuss mailing list > zfs-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Roch Bourbonnais wrote:> Le 12 janv. 09 ? 17:39, Carson Gaspar a ?crit : > > >> Joerg Schilling wrote: >> >>> Fabian W?rner <fabian.woerner at googlemail.com> wrote: >>> >>> >>>> my post was not to start a discuss gpl<>cddl. >>>> It just an idea to promote ZFS and OPENSOLARIS!!!! >>>> If it was against anything than against exfat, nothing else!!! >>>> >>> If you like to promoote ZFS, you need to understand why the party >>> you like >>> to promote it to does not already use it ;-) >>> >> And for SDXC, ZFS will probably never be the filesystem of choice. >> Removable media of this type is mostly used in portable electronic >> devices, such as cameras, cellphones, etc. All of which are power, >> CPU, >> and memory limited. ZFS, while a marvelous filesystem, is incredibly >> RAM >> hungry. I suspect it''s CPU profile is also non-trivial for a >> restricted >> performance device. >> > > I have not looked at it recently but for any access greater than ~ 16K > ZFS was more efficient than UFS. > It''s just one partial data point but the conventional wisdom that ZFS > will use more cpu is not an absolute truth. > > Even more so for RAM, ZFS with 128K record make efficient use of > metadata. The only ram it needs to operation is 10 seconds of > of your workload''s throughput and that can be tuned down in appliances. >DOS/FAT filesystem implementations in appliances can be found in less than 8K code and data size (mostly that''s code). Limited functionality implementations can be smaller than 1kB size. -- Andrew
On Tue, Jan 13, 2009 at 6:26 AM, Brian Wilson <bfwilson at doit.wisc.edu>wrote:> > Does creating ZFS pools on multiple partitions on the same physical drive > still run into the performance and other issues that putting pools in slices > does? >Is zfs going to own the whole drive or not? The *issue* is that zfs will not use the drive cache if it doesn''t own the whole disk since it won''t know whether or not it should be flushing cache at any given point in time. It could cause corruption if you had UFS and zfs on the same disk. --Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090114/2cadf6d6/attachment.html>
On Wed, Jan 14, 2009 at 20:03, Tim <tim at tcsac.net> wrote:> > > On Tue, Jan 13, 2009 at 6:26 AM, Brian Wilson <bfwilson at doit.wisc.edu> > wrote: >> >> Does creating ZFS pools on multiple partitions on the same physical drive >> still run into the performance and other issues that putting pools in slices >> does? > > > Is zfs going to own the whole drive or not? The *issue* is that zfs will > not use the drive cache if it doesn''t own the whole disk since it won''t know > whether or not it should be flushing cache at any given point in time.ZFS will always flush the disk cache at appropriate times. If ZFS thinks that is alone it will turn the write cache on the disk on.> > It could cause corruption if you had UFS and zfs on the same disk.It is safe to have UFS and ZFS on the same disk and it has always been safe. Write cache on the disk is not safe for UFS, that is why zfs will turn it on only if it is alone.
but, the write cache on/offness is a stateful setting stored on the disk platter, right? so it survives reboots of the disk, and ZFS doesn''t turn it off, and UFS arguably should turn it off but doesn''t---once you''ve dedicated a disk to ZFS, you have to turn the write cache off yourself somehow using ''format -e'' if you are no longer using a disk for ZFS only. Or am I remembering wrong? -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 304 bytes Desc: not available URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090114/41b5c5fc/attachment.bin>
On Wed, Jan 14, 2009 at 2:48 PM, Miles Nordin <carton at ivy.net> wrote:> but, the write cache on/offness is a stateful setting stored on the > disk platter, right? so it survives reboots of the disk, and ZFS > doesn''t turn it off, and UFS arguably should turn it off but > doesn''t---once you''ve dedicated a disk to ZFS, you have to turn the > write cache off yourself somehow using ''format -e'' if you are no > longer using a disk for ZFS only. Or am I remembering wrong? >ZFS does turn it off if it doesn''t have the whole disk. That''s where the performance issues come from. --Tim -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090114/bb73cd56/attachment.html>
On Wed, Jan 14, 2009 at 2:40 PM, Mattias Pantzare <pantzare at gmail.com>wrote:> On Wed, Jan 14, 2009 at 20:03, Tim <tim at tcsac.net> wrote: > > > > > > On Tue, Jan 13, 2009 at 6:26 AM, Brian Wilson <bfwilson at doit.wisc.edu> > > wrote: > >> > >> Does creating ZFS pools on multiple partitions on the same physical > drive > >> still run into the performance and other issues that putting pools in > slices > >> does? > > > > > > Is zfs going to own the whole drive or not? The *issue* is that zfs will > > not use the drive cache if it doesn''t own the whole disk since it won''t > know > > whether or not it should be flushing cache at any given point in time. > > ZFS will always flush the disk cache at appropriate times. If ZFS > thinks that is alone it will turn the write cache on the disk on.I''m not sure if you''re trying to argue or agree. If you''re trying to argue, you''re going to have to do a better job than "zfs will always flush disk cache at appropriate times", because that''s outright false in the case where zfs doesn''t own the entire disk. That flush may very well produce an outcome zfs could never pre-determine.> > > It could cause corruption if you had UFS and zfs on the same disk. > > It is safe to have UFS and ZFS on the same disk and it has always been > safe. >***unless you turn on write cache. And without write cache, performance sucks. Hence me answering the OP''s question.> > Write cache on the disk is not safe for UFS, that is why zfs will turn > it on only if it is alone.Which is EXACTLY what he''s asking, and what I just told him. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090114/b6d254f5/attachment.html>
?? Tim wrote:> On Wed, Jan 14, 2009 at 2:40 PM, Mattias Pantzare <pantzare at gmail.com > <mailto:pantzare at gmail.com>> wrote: > > On Wed, Jan 14, 2009 at 20:03, Tim <tim at tcsac.net > <mailto:tim at tcsac.net>> wrote: > > > > > > On Tue, Jan 13, 2009 at 6:26 AM, Brian Wilson > <bfwilson at doit.wisc.edu <mailto:bfwilson at doit.wisc.edu>> > > wrote: > >> > >> Does creating ZFS pools on multiple partitions on the same > physical drive > >> still run into the performance and other issues that putting > pools in slices > >> does? > > > > > > Is zfs going to own the whole drive or not? The *issue* is that > zfs will > > not use the drive cache if it doesn''t own the whole disk since it > won''t know > > whether or not it should be flushing cache at any given point in > time. > > ZFS will always flush the disk cache at appropriate times. If ZFS > thinks that is alone it will turn the write cache on the disk on. > > > I''m not sure if you''re trying to argue or agree. If you''re trying to > argue, you''re going to have to do a better job than "zfs will always > flush disk cache at appropriate times", because that''s outright false in > the case where zfs doesn''t own the entire disk. That flush may very > well produce an outcome zfs could never pre-determine.Would you care to explain this logic? Are you saying that if ZFS sends a cache flush command to a disk that it will "produce an outcome ZFS could never pre-determime?" Or am I just misinterpreting? -- richard
>> >> ZFS will always flush the disk cache at appropriate times. If ZFS >> thinks that is alone it will turn the write cache on the disk on. > > I''m not sure if you''re trying to argue or agree. If you''re trying to argue, > you''re going to have to do a better job than "zfs will always flush disk > cache at appropriate times", because that''s outright false in the case where > zfs doesn''t own the entire disk. That flush may very well produce an > outcome zfs could never pre-determine.You can send flush cache commands to the disk how often you wish, the only thing that happens is that the disk writes dirty sectors from its cache to the disk. That is, no writes will be done that should not have happend at some time anyway. This will not harm UFS or any other user of the disk. Other users can issue flush cache command without affecting ZFS. Please read up on what the flus cache command does! ZFS will send flush cache commands even when it is not alone on the disk. There are many disks with write cache on by default. There have even been disks that won''t turn it off even if told so.>> > It could cause corruption if you had UFS and zfs on the same disk. >> >> It is safe to have UFS and ZFS on the same disk and it has always been >> safe. > > ***unless you turn on write cache. And without write cache, performance > sucks. Hence me answering the OP''s question.There was no mention of cache at all in the question. It was not clear that this sentence reffered to your own text, hence the misunderstanding: "It could cause corruption if you had UFS and zfs on the same disk." I read that as a separate statement. As to the performance sucks, that is putting it a bit harsh, you will get better performance with write cache but the system will be perfectly usable without.
On 14 Jan 2009, at 10:01, Andrew Gabriel wrote:> DOS/FAT filesystem implementations in appliances can be found in less > than 8K code and data size (mostly that''s code). Limited functionality > implementations can be smaller than 1kB size.Just for the sake of comparison, how big is the limited ZFS implementation in grub? Cheers, Chris
>ZFS does turn it off if it doesn''t have the whole disk. That''s where the >performance issues come from.But it doesn''t "touch it" so ZFS continues to work if you enable write caching. And I think we default to "write-cache" enabled for ATA/IDE disks. (The reason is that they''re shipped with write cache enabled and that''s the only setting tested by the manufacturer) Casper
For the sake of curiosity, is it safe to have components of two different ZFS pools on the same drive, with and without HDD write cache turned on? How will ZFS itself behave, would it turn on the disk cache if the two imported pools co-own the drive? An example is a multi-disk system like mine which had UFS-mirrored root and a ZFS-raidz2 for data, then was upgraded to a ZFS-mirrored root with another ZFS pool for data (partially since grub doesn''t do ZFS roots on zfs-raidz*). PS: for a system with 1 or 2 drives, is there any preference to either layout: 1) sliced with 2 zpools (root and data) 2) sliced with one zpool for all 3) whole-disk with 1 zpool for all (is that supported by GRUB for boot disks at all?) -- This message posted from opensolaris.org
Tim writes: > On Tue, Jan 13, 2009 at 6:26 AM, Brian Wilson <bfwilson at doit.wisc.edu>wrote: > > > > > Does creating ZFS pools on multiple partitions on the same physical drive > > still run into the performance and other issues that putting pools in slices > > does? > > > > > Is zfs going to own the whole drive or not? The *issue* is that zfs will > not use the drive cache if it doesn''t own the whole disk since it won''t know > whether or not it should be flushing cache at any given point in time. > > It could cause corruption if you had UFS and zfs on the same disk. > Let me correct a few things. ZFS unconditionaly flushes the write caches when it needs to and owning a drive or not is not important for the consistency of ZFS. If ZFS owns a disk it will enable the write cache on the drive but I''m not positive this has a great performance impact today. It used to but that was before we had a proper NCQ implementation. Today I don''t know that it helps much. That this is because we always flush the cache when consistency requires it. The performance issue of using a drive to multiple unrelated consumers (ZFS & UFS) is that, if both are active at the same time, this will defeat the I/O scheduling smarts implemented in ZFS. Rather than have data streaming to some physical location of the rust, the competition of UFS for I/O will cause extra head movement. -r
>The performance issue of using a drive to multiple unrelated >consumers (ZFS & UFS) is that, if both are active at the >same time, this will defeat the I/O scheduling smarts >implemented in ZFS. Rather than have data streaming to some >physical location of the rust, the competition of UFS for >I/O will cause extra head movement.Solaris still makes sure that blocks are sorted, whether they come from UFS or from ZFS. Casper
On Thu, Jan 15, 2009 at 21:51, <Casper.Dik at sun.com> wrote:> > >>The performance issue of using a drive to multiple unrelated >>consumers (ZFS & UFS) is that, if both are active at the >>same time, this will defeat the I/O scheduling smarts >>implemented in ZFS. Rather than have data streaming to some >>physical location of the rust, the competition of UFS for >>I/O will cause extra head movement. > > > Solaris still makes sure that blocks are sorted, whether they > come from UFS or from ZFS.Yes, but consider the common case where UFS and ZFS are on separate slices of the disk. Then writes that ZFS thinks will be contiguous aren''t, because the disk has seeked (sought?) to the UFS slice, a long way away. Solaris will optimize this as much as possible, sure, but there''s nothing you can do to avoid moving the head back and forth from one slice to the other. There''s only so far you can sort blocks 12 and 10**7 ;) Will
On Mon, Jan 12 at 10:00, Casper.Dik at sun.com wrote:>>My impression is not that other OS''s aren''t interested in ZFS, they >>are, it''s that the licensing restrictions limit native support to >>Solaris, BSD, and OS-X. >> >>If you wanted native support in Windows or Linux, it would require a >>significant effort from Sun. > > >Why is that a problem for Windows? Linux, yes, but if they want they can >change that.Who is "they" ? It''s not a problem, it just is-what-it-is. The "significant effort" I am referring to is changes to the licensing, which is a tricky endeavour as soon as you have contributors instead of a contributor. Doesn''t really matter who changes, or really if anyone changes at all. --eric -- Eric D. Mudama edmudama at mail.bounceswoosh.org