[Ocfs2-users] Re: Ocfs2-users Digest, Vol 41, Issue 21
Arjan Widlak
arjan at unitedknowledge.nl
Tue May 22 14:34:02 PDT 2007
Dear all,
> Caveats
> =======
> Features which OCFS2 does not support yet:
> - extended attributes
> - readonly mount
> - shared writeable mmap
> - loopback is supported, but data written will not
> be cluster coherent.
> - quotas
> - cluster aware flock <----------------------------------------
> - Directory change notification (F_NOTIFY)
> - Distributed Caching (F_SETLEASE/F_GETLEASE/break_lease)
> - POSIX ACLs
> - readpages / writepages (not user visible)
If this is an ill-informed question, please enlighten me. I'm following
this list for quite some time and have done some tests (quite some time
ago) and noticed then that if you write from two machines to the same
file at the same moment, the one that is finished last 'wins' i.c. the
file contains the content that that machine tried to write.
I'm reading this list to see when this issue is resolved. Now I'm
reading this, I thought: let's ask. If cluster aware file-locking is not
supported, what exactly is the use of a cluster file system? Isn't that
the most important feature?
Kind regards,
Arjan.
> Caveats
> =======
> Features which OCFS2 does not support yet:
> - extended attributes
> - readonly mount
> - shared writeable mmap
> - loopback is supported, but data written will not
> be cluster coherent.
> - quotas
> - cluster aware flock <----------------------------------------
> - Directory change notification (F_NOTIFY)
> - Distributed Caching (F_SETLEASE/F_GETLEASE/break_lease)
> - POSIX ACLs
> - readpages / writepages (not user visible)
On Mon, 2007-05-21 at 14:00 -0500, ocfs2-users-request at oss.oracle.com
wrote:
> Send Ocfs2-users mailing list submissions to
> ocfs2-users at oss.oracle.com
>
> To subscribe or unsubscribe via the World Wide Web, visit
> http://oss.oracle.com/mailman/listinfo/ocfs2-users
> or, via email, send a message with subject or body 'help' to
> ocfs2-users-request at oss.oracle.com
>
> You can reach the person managing the list at
> ocfs2-users-owner at oss.oracle.com
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Ocfs2-users digest..."
>
>
> Today's Topics:
>
> 1. php flock function and ocfs2? (Stephan Hendl)
> 2. Re: php flock function and ocfs2? (Stephan Hendl)
> 3. slow file creation (Matthew Flusche)
> 4. Re: slow file creation (Sunil Mushran)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Mon, 21 May 2007 15:30:50 +0200
> From: "Stephan Hendl" <Stephan.Hendl at lds.brandenburg.de>
> Subject: [Ocfs2-users] php flock function and ocfs2?
> To: <ocfs2-users at oss.oracle.com>
> Message-ID: <4651BB2A.4F9C.0016.0 at lds.brandenburg.de>
> Content-Type: text/plain; charset=US-ASCII
>
> Hi all,
>
> does anybody have experiences with php flock funktion and ocfs2 clusters? flock is described in http://php.net/flock . The following text is outlined by the developers:
>
> -------- snip --------
> flock() will not work on NFS and many other networked file systems. Check your operating system documentation for more details.
>
> On some operating systems flock() is implemented at the process level. When using a multithreaded server API like ISAPI you may not be able to rely on flock() to protect files against other PHP scripts running in parallel threads of the same server instance!
> -------- snip --------
>
> OS is CentOS 4.4, latest kernel (2.6.9-42.0.10ELsmp) and latest rpm package from ocfs2.
>
> Could it be that there are interactions between the two locking mechanism of php and ocfs2?
>
> Stephan
>
>
>
>
> ------------------------------
>
> Message: 2
> Date: Mon, 21 May 2007 16:20:02 +0200
> From: "Stephan Hendl" <Stephan.Hendl at lds.brandenburg.de>
> Subject: Re: [Ocfs2-users] php flock function and ocfs2?
> To: <ocfs2-users at oss.oracle.com>
> Message-ID: <4651C6B1.4F9C.0016.0 at lds.brandenburg.de>
> Content-Type: text/plain; charset=US-ASCII
>
> in the cluster docu I found:
>
> Caveats
> =======
> Features which OCFS2 does not support yet:
> - extended attributes
> - readonly mount
> - shared writeable mmap
> - loopback is supported, but data written will not
> be cluster coherent.
> - quotas
> - cluster aware flock <----------------------------------------
> - Directory change notification (F_NOTIFY)
> - Distributed Caching (F_SETLEASE/F_GETLEASE/break_lease)
> - POSIX ACLs
> - readpages / writepages (not user visible)
>
> Could this be the reason and when yes will it be implemented in the near future?
>
> Stephan
>
> >>> "Stephan Hendl" <Stephan.Hendl at lds.brandenburg.de> schrieb am 21.05.2007 um
> 15:30 in Nachricht <4651BB2A.4F9C.0016.0 at lds.brandenburg.de>:
> > Hi all,
> >
> > does anybody have experiences with php flock funktion and ocfs2 clusters?
> > flock is described in http://php.net/flock . The following text is outlined
> > by the developers:
> >
> > -------- snip --------
> > flock() will not work on NFS and many other networked file systems. Check
> > your operating system documentation for more details.
> >
> > On some operating systems flock() is implemented at the process level. When
> > using a multithreaded server API like ISAPI you may not be able to rely on
> > flock() to protect files against other PHP scripts running in parallel
> > threads of the same server instance!
> > -------- snip --------
> >
> > OS is CentOS 4.4, latest kernel (2.6.9-42.0.10ELsmp) and latest rpm package
> > from ocfs2.
> >
> > Could it be that there are interactions between the two locking mechanism of
> > php and ocfs2?
> >
> > Stephan
> >
> >
> > _______________________________________________
> > Ocfs2-users mailing list
> > Ocfs2-users at oss.oracle.com
> > http://oss.oracle.com/mailman/listinfo/ocfs2-users
>
>
>
>
> ------------------------------
>
> Message: 3
> Date: Mon, 21 May 2007 11:32:58 -0500
> From: "Matthew Flusche" <Matthew.Flusche at tvguide.com>
> Subject: [Ocfs2-users] slow file creation
> To: <ocfs2-users at oss.oracle.com>
> Message-ID:
> <7A6678D6C264194DA3D2C549536075AD0134B45B at tul1mail1.tvguide.com>
> Content-Type: text/plain; charset="us-ascii"
>
> Hi all,
>
> I'm troubleshooting an ocfs2 performance problem where creating files in
> a directory containing ~180k files is taking significant time to
> complete. Sometimes creating an empty file will take >100 seconds to
> complete. This is a three node cluster. I'm currently running OCFS2
> 1.2.3-1. Are there any changes in a recent version that may address
> this issue? What should I look at to troubleshoot? The systems overall
> utilization is low.
>
> (All 3 systems are identical)
> RHEL 4 update 4 - 2.6.9-42.0.2.ELhugemem (32-bit)
> ocfs2-2.6.9-42.0.2.ELhugemem-1.2.3-1
> ocfs2-tools-1.2.2-1
>
> HP Proliant DL385 - 2 X 2.4 GHz AMD Opteron Processor 250
> 16GB RAM
> 2 X emulex LP10000 HBAs
> Powerpath 4.5.1
>
> shared disks are from an EMC clariion CX500
>
>
> Thanks,
>
> Matt
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: http://oss.oracle.com/pipermail/ocfs2-users/attachments/20070521/c868d69c/attachment-0001.html
>
> ------------------------------
>
> Message: 4
> Date: Mon, 21 May 2007 09:39:59 -0700
> From: Sunil Mushran <Sunil.Mushran at oracle.com>
> Subject: Re: [Ocfs2-users] slow file creation
> To: Matthew Flusche <Matthew.Flusche at tvguide.com>
> Cc: ocfs2-users at oss.oracle.com
> Message-ID: <4651CB5F.7070509 at oracle.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> 1. Upgrade to atleast 1.2.4 if not 1.2.5. We fixed a lowmem
> issue in 1.2.4 that reduces the fs' memory footprint.
>
> 2. strace -tt -T the process. That will narrow down the problem
> area.
>
> Matthew Flusche wrote:
> > Hi all,
> >
> > I'm troubleshooting an ocfs2 performance problem where creating files
> > in a directory containing ~180k files is taking significant time to
> > complete. Sometimes creating an empty file will take >100 seconds to
> > complete. This is a three node cluster. I'm currently running OCFS2
> > 1.2.3-1. Are there any changes in a recent version that may address
> > this issue? What should I look at to troubleshoot? The systems
> > overall utilization is low.
> >
> > (All 3 systems are identical)
> > RHEL 4 update 4 - 2.6.9-42.0.2.ELhugemem (32-bit)
> > ocfs2-2.6.9-42.0.2.ELhugemem-1.2.3-1
> > ocfs2-tools-1.2.2-1
> >
> > HP Proliant DL385 - 2 X 2.4 GHz AMD Opteron Processor 250
> > 16GB RAM
> > 2 X emulex LP10000 HBAs
> > Powerpath 4.5.1
> >
> > shared disks are from an EMC clariion CX500
> >
> >
> > Thanks,
> >
> > Matt
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > Ocfs2-users mailing list
> > Ocfs2-users at oss.oracle.com
> > http://oss.oracle.com/mailman/listinfo/ocfs2-users
>
>
>
>
> ------------------------------
>
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users
>
> End of Ocfs2-users Digest, Vol 41, Issue 21
> *******************************************
--
Arjan Widlak <arjan at unitedknowledge.nl>
United Knowledge | Politiek-digitaal.nl
More information about the Ocfs2-users
mailing list