You may/should get better performance with '-r 4m' since that is the
default amount of buffer space allocated for each I/O request.

Dean

On Thu, 1 Dec 2005, Rob Ross wrote:

> FYI, more information from Praveen.
>
> Rob
>
> -------- Original Message --------
> Subject: Re: [PVFS2-users] pvfs2 scaling with addition of data servers
> Date: Tue, 22 Nov 2005 14:24:17 -0800
> From: Praveen KJ <[EMAIL PROTECTED]>
> To: Rob Ross <[EMAIL PROTECTED]>
> References: <[EMAIL PROTECTED]> <[EMAIL PROTECTED]>
>
> Rob,
>         Thanks for the prompt response.  I am using iozone on a single
> client to read the file.
> The throughput is reported by iozone. There is nothing happening on the
> system. The whole system
> including client and servers are totally homogenous and run Rocks 4.1
>
>         Iozone command line looks like ..   'iozone -r 1m -w -s 5g -i 1'
> Let me know if I have to tweak something in the system. Like the stripe
> size etc...
> We have the ability to experiment fast to try out different options.
>
>         Also when I start the system, *df -h* reports 9%-14% of pvfs
> partition space ( depending on number of servers)   as filled.
> This is with zero files to begin with.  Is this the usual overhead ?
>
>
> Thanks,
> Praveen
>
> Rob Ross wrote:
>
> > Hi,
> >
> > Can you tell us more about what you're doing?  How are you reading the
> > file?  How are you measuring the performance?  Is anything else
> > happening on the system?
> >
> > Thanks,
> >
> > Rob
> >
> > Praveen KJ wrote:
> >
> >> Hi,
> >>     We have a pvfs2 client reading a 5 GB file sequentially from
> >> multiple pvfs2 data servers.
> >>     There is just one metadata server. All the nodes are connected
> >> via GigE network.
> >>
> >>     The problem we are seeing is that the single-client read
> >> throughput decreases as the number of servers increase.
> >>     With 24 servers we see around 5 MB/s read compared to 50 MB/s
> >> read with just 2 servers.
> >>     The setup between each experiment is exactly same, except the
> >> number of servers.
> >>
> >>     We have tried a couple of pvfs2 versions including the latest 1.3.1.
> >>     Is there any setting we have to tweak ?  We have taken only pvfs2
> >> defaults till now.
> >>     Some of the relevant options are pasted below.
> >>
> >> Thanks,
> >> Praveen
> >>
> >> ------------------------------------------
> >> pvfs2-fs.conf ( except hostnames )
> >> ------------------------------------------
> >> <Defaults>
> >>       UnexpectedRequests 50
> >>       LogFile /tmp/pvfs2-server.log
> >>       EventLogging none
> >>       LogStamp usec
> >>       BMIModules bmi_tcp
> >>       FlowModules flowproto_multiqueue
> >>       PerfUpdateInterval 1000
> >>       ServerJobBMITimeoutSecs 30
> >>       ServerJobFlowTimeoutSecs 30
> >>       ClientJobBMITimeoutSecs 300
> >>       ClientJobFlowTimeoutSecs 300
> >>       ClientRetryLimit 5
> >>       ClientRetryDelayMilliSecs 2000
> >> </Defaults>
> >>
> >> <Filesystem>
> >>       Name pvfs2-fs
> >>       ID 1175641402
> >>       RootHandle 1048576
> >>       <StorageHints>
> >>               TroveSyncMeta yes
> >>               TroveSyncData no
> >>               AttrCacheKeywords datafile_handles,metafile_dist
> >>               AttrCacheKeywords dir_ent, symlink_target
> >>               AttrCacheSize 4093
> >>               AttrCacheMaxNumElems 32768
> >>       </StorageHints>
> >> </Filesystem>
> >> _______________________________________________
> >> PVFS2-users mailing list
> >> [email protected]
> >> http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users
> >>
>
> _______________________________________________
> PVFS2-users mailing list
> [email protected]
> http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users
>
_______________________________________________
PVFS2-users mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users

Reply via email to