[zfs-discuss] ZFS and dual-pathed A5200
Hi, I understand that ZFS leaves multipathing to MPXIO or the like. For a combination of dual-path A5200 with QLGC2100 HBAs (non-leadville stack), how would ZFS react in seeing this ? WL ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] legato support
Gregory Shaw wrote: Hey, does anybody know the timeframe for when Legato Networker will support ZFS? Looks like EMC has released Networker version 7.3.2 which has ZFS support. -Mark ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and IBM sdd (vpath)
You are right I am not the first one, with a little bit a research I found this tread: http://mail.opensolaris.org/pipermail/zfs-discuss/2006-July/003937.html > On Wed, Oct 18, 2006 at 06:57:21AM -0700, Geoffroy Doucet wrote: >> Hello, I am trying to configure ZFS with IBM sdd. IBM sdd is like >> powerpath, MPXIO or VxDMP. >> >> Here is the error message when I try to create my pool: >> bash-3.00# zpool create tank /dev/dsk/vpath1a >> warning: device in use checking failed: No such device >> internal error: unexpected error 22 at line 446 of >> ../common/libzfs_pool.c >> bash-3.00# zpool create tank /dev/dsk/vpath1c >> cannot open '/dev/dsk/vpath1c': I/O error >> bash-3.00# zpool create tank vpath1 >> cannot open 'vpath1': no such device in /dev/dsk >> must be a full path or shorthand device name >> bash-3.00# zpool create tank vpath1c >> cannot open '/dev/dsk/vpath1c': I/O error >> bash-3.00# zpool create tank vpath1a >> warning: device in use checking failed: No such device >> internal error: unexpected error 22 at line 446 of >> ../common/libzfs_pool.c > > My guess for the first case (vpath1a) is that the IBM driver is not > correctly implementing the necessary DDI properties for ZFS to determine > the configuration of the device. What bits are you using? Off the top > of my head, this should catch the most likely candidate: > > # dtrace -n ldi_get_size:return'{trace(arg1)}' > > I'd also check to see if there are any updates from IBM. From the looks > of it, 'vpath1c' cannot even be opened from userland, so something else > seems misconfigured in that case. > > - Eric > > -- > Eric Schrock, Solaris Kernel Development > http://blogs.sun.com/eschrock > ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and IBM sdd (vpath)
Eric Schrock wrote: Chances are they have already heard of this bug, as I seem to remember it coming up before. Comes up all the time on the storage lists. Especially, when people try to get mpxio to work with the underlying arrays instead of sdd(vpath). Last I recall they aren't exporting the LUNs with the proper IEEE Registered Extended bitsbut that was awhile ago. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and IBM sdd (vpath)
On Wed, Oct 18, 2006 at 02:15:42PM -0400, Geoffroy Doucet wrote: > Here is the ouput of your DTrace script: > > # dtrace -n ldi_get_size:return'{trace(arg1)}' > dtrace: description 'ldi_get_size:return' matched 1 probe > CPU IDFUNCTION:NAME > 1 24930 ldi_get_size:return-1 > > With the command: > # zpool create tank /dev/dsk/vpath1c > warning: device in use checking failed: No such device > internal error: unexpected error 22 at line 446 of ../common/libzfs_pool.c Yep. This is a bug in the IBM driver code. There are a number of ddi properties ("NBlocks", "nblocks", "Size", and "size") that a driver can export we can calculate the size. Looks like the IBM driver is exporting none of these, so we have no (generic) way of knowing how big the underlying device is. I would escalate this with IBM, and point to the ldi_get_size() source on opensolaris.org for information on what we expect. Chances are they have already heard of this bug, as I seem to remember it coming up before. - Eric -- Eric Schrock, Solaris Kernel Development http://blogs.sun.com/eschrock ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and IBM sdd (vpath)
Here is the ouput of your DTrace script: # dtrace -n ldi_get_size:return'{trace(arg1)}' dtrace: description 'ldi_get_size:return' matched 1 probe CPU IDFUNCTION:NAME 1 24930 ldi_get_size:return-1 With the command: # zpool create tank /dev/dsk/vpath1c warning: device in use checking failed: No such device internal error: unexpected error 22 at line 446 of ../common/libzfs_pool.c > On Wed, Oct 18, 2006 at 06:57:21AM -0700, Geoffroy Doucet wrote: >> Hello, I am trying to configure ZFS with IBM sdd. IBM sdd is like >> powerpath, MPXIO or VxDMP. >> >> Here is the error message when I try to create my pool: >> bash-3.00# zpool create tank /dev/dsk/vpath1a >> warning: device in use checking failed: No such device >> internal error: unexpected error 22 at line 446 of >> ../common/libzfs_pool.c >> bash-3.00# zpool create tank /dev/dsk/vpath1c >> cannot open '/dev/dsk/vpath1c': I/O error >> bash-3.00# zpool create tank vpath1 >> cannot open 'vpath1': no such device in /dev/dsk >> must be a full path or shorthand device name >> bash-3.00# zpool create tank vpath1c >> cannot open '/dev/dsk/vpath1c': I/O error >> bash-3.00# zpool create tank vpath1a >> warning: device in use checking failed: No such device >> internal error: unexpected error 22 at line 446 of >> ../common/libzfs_pool.c > > My guess for the first case (vpath1a) is that the IBM driver is not > correctly implementing the necessary DDI properties for ZFS to determine > the configuration of the device. What bits are you using? Off the top > of my head, this should catch the most likely candidate: > > # dtrace -n ldi_get_size:return'{trace(arg1)}' > > I'd also check to see if there are any updates from IBM. From the looks > of it, 'vpath1c' cannot even be opened from userland, so something else > seems misconfigured in that case. > > - Eric > > -- > Eric Schrock, Solaris Kernel Development > http://blogs.sun.com/eschrock > ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Re: Configuring a 3510 for ZFS
Robert Milkowski wrote: Well, I don't know - that way you end doing RAID in ZFS anyway so probably doing just RAID-10 in ZFS without ditto block would be better. The win with ditto blocks is allowing you to recover from a data inconsistency at the fs level as opposed to dealing with a block error within the raid group. The problem, as discussed previously, is in the accounting and management of the ditto blocks. ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Zfs Performance with millions of small files in Sendmail messaging environment]
The zap is really speedy when it comes to file lookups in a directory. If you're really concerned about performance, and depending on your system you're running on, I'd probably recommend staying below 10 million or so files per directory. Just since you'll get fastest results when we can fit most of zap blocks in the ARC as opposed to having to go and fetch them from disk. *(disclaimer: this was a benchmark i ran on bits from February on an opteron box with 3 GB of memory with a single scsi disk in the pool. So you may want to rerun with current bits for better numbers if this is very important to you expect to be storing millions of files per directory on a regular basis). You can also go nuts on creating filesystems. I believe the only thing to be forewarned of is that upon reboot, you'll be mounting all these filesystems so this might slow down the reboot process somewhat when you get to *many thousands(due to the way the automounter behaves). I believe that a fix is being worked on (may already be back? ) to make this process much quicker, however I'm not sure. Anybody know what the status of that is? Noel On Oct 15, 2006, at 11:08 PM, Ramneek Sethi wrote: Thanks for your feedback.. any comments on VxFS.. Also is there any limit of files within filesystem or within directory for optimal performance.. Thanks Robert Milkowski wrote On 10/14/06 04:19,: Hello Ramneek, Friday, October 13, 2006, 6:07:22 PM, you wrote: RS> Hello Experts RS> Would appreciate if somebody can comment on sendmail environment on RS> solaris 10. RS> How will Zfs perform if one has millions of files in sendmail message RS> store directory under zfs filesystem compared to UFS or VxFS.. Actually not sendmail but also MTA and ZFS is about 5% better in real production than UFS+SVM. -- Thanks & Regards, *** _/_/_/ _/_/ _/ _/ Ramneek Sethi _/ _/_/ _/_/ _/ Systems Support Engineer _/_/_/ _/_/ _/ _/ _/Sun Microsystems India Pvt. Ltd. _/ _/_/ _/ _/_/ 5th Floor,Right Wing , _/_/_/ _/_/_/ _/ _/ The Capital Court,Munirka New Delhi - 110067,INDIA Phone : 91--11-42191029 Fax : 91-11-26160928 Support SERVICESE-mail : [EMAIL PROTECTED] *** For any Support Queries pls dial the Toll Free Number 1600-4254-786 *** ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS and IBM sdd (vpath)
On Wed, Oct 18, 2006 at 06:57:21AM -0700, Geoffroy Doucet wrote: > Hello, I am trying to configure ZFS with IBM sdd. IBM sdd is like powerpath, > MPXIO or VxDMP. > > Here is the error message when I try to create my pool: > bash-3.00# zpool create tank /dev/dsk/vpath1a > warning: device in use checking failed: No such device > internal error: unexpected error 22 at line 446 of ../common/libzfs_pool.c > bash-3.00# zpool create tank /dev/dsk/vpath1c > cannot open '/dev/dsk/vpath1c': I/O error > bash-3.00# zpool create tank vpath1 > cannot open 'vpath1': no such device in /dev/dsk > must be a full path or shorthand device name > bash-3.00# zpool create tank vpath1c > cannot open '/dev/dsk/vpath1c': I/O error > bash-3.00# zpool create tank vpath1a > warning: device in use checking failed: No such device > internal error: unexpected error 22 at line 446 of ../common/libzfs_pool.c My guess for the first case (vpath1a) is that the IBM driver is not correctly implementing the necessary DDI properties for ZFS to determine the configuration of the device. What bits are you using? Off the top of my head, this should catch the most likely candidate: # dtrace -n ldi_get_size:return'{trace(arg1)}' I'd also check to see if there are any updates from IBM. From the looks of it, 'vpath1c' cannot even be opened from userland, so something else seems misconfigured in that case. - Eric -- Eric Schrock, Solaris Kernel Development http://blogs.sun.com/eschrock ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS, home and Linux
msl wrote: Hello, I'm trying to implement a NAS server with solaris/NFS and, of course, ZFS. But for that, we have a little problem... what about the /home filesystem? I mean, i have a lot of linux clients, and the "/home" directory is on a NFS server (today, linux). I want to use ZFS, and change the "directory" home like /home/leal, to "filesystems" like /home/leal (just like the documentation recommends). Now, a PAM module (pam_mkhomedir) have solved that problem, creating the user home directory (as demand). Do you have a solution to that? Like a linux PAM module to create a zfs filesystem under /home, and not a ordinary directory? So i can have a per user filesystem under "/home/"... Of course, with a service on the NFS solaris server... I'm asking because if you agree that it is a "problem", we can create a project in opensolaris to work on it. Or maybe, you have a trivial solution, that i'm not seeing. Thanks very much for your time! Have an executable automounter map that runs on the NFS server do that for you. See automount(1M) for information on how to write the maps. You could also port the pam_mkhomedir module and have it do a zfs create instead of mkdir. PAM after all started on Solaris :-) -- Darren J Moffat ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] ZFS and IBM sdd (vpath)
Hello, I am trying to configure ZFS with IBM sdd. IBM sdd is like powerpath, MPXIO or VxDMP. Here is the error message when I try to create my pool: bash-3.00# zpool create tank /dev/dsk/vpath1a warning: device in use checking failed: No such device internal error: unexpected error 22 at line 446 of ../common/libzfs_pool.c bash-3.00# zpool create tank /dev/dsk/vpath1c cannot open '/dev/dsk/vpath1c': I/O error bash-3.00# zpool create tank vpath1 cannot open 'vpath1': no such device in /dev/dsk must be a full path or shorthand device name bash-3.00# zpool create tank vpath1c cannot open '/dev/dsk/vpath1c': I/O error bash-3.00# zpool create tank vpath1a warning: device in use checking failed: No such device internal error: unexpected error 22 at line 446 of ../common/libzfs_pool.c Is there anyone else that use ZFS with IBM sdd? This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
RE: [zfs-discuss] Re: Configuring a 3510 for ZFS
These are pretty much the conclusions we reached from this discussion, and thanks for all the input. On the 3510 we are configuring 12 nraid LUNs - basically presenting the 12 disks to the OS as they are. In a real scenario we will be mirroring across two 3510s anyway. We have also decided against raidz in this scenario. Regards, Ciaran. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Anantha N. Srirama Sent: 18 October 2006 13:11 To: zfs-discuss@opensolaris.org Subject: [zfs-discuss] Re: Configuring a 3510 for ZFS Thanks for the stimulating exchange of ideas/thoughts. I've always been a believer of letting s/w do my RAID functions; for example in the old days of VxVM I always preferred to do mirroring at the s/w level. It is my belief that there is more 'meta' information available at the OS level than at the storage level for s/w to make intelligent decisions; dynamic recordsize in ZFS is one example. Any thoughts on the following approach: 1. I'll configure 3511 to present multiple LUNs (mirrored internally) to OS. 2. Lay down a ZFS pool/filesystem without RAID protection (RAIDZ...) in the OS With this approach I will enjoy the caching facility of 3511 and the checksum protection afforded by ZFS. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Re: Configuring a 3510 for ZFS
Thanks for the stimulating exchange of ideas/thoughts. I've always been a believer of letting s/w do my RAID functions; for example in the old days of VxVM I always preferred to do mirroring at the s/w level. It is my belief that there is more 'meta' information available at the OS level than at the storage level for s/w to make intelligent decisions; dynamic recordsize in ZFS is one example. Any thoughts on the following approach: 1. I'll configure 3511 to present multiple LUNs (mirrored internally) to OS. 2. Lay down a ZFS pool/filesystem without RAID protection (RAIDZ...) in the OS With this approach I will enjoy the caching facility of 3511 and the checksum protection afforded by ZFS. This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Re: ZFS Inexpensive SATA Whitebox
Richard Elling - PAE schrieb: Where most people get confused is the expectation that a hot-plug device works like a hot-swap device. Well, seems like you should also inform your documentation team about this definition: http://www.sun.com/products-n-solutions/hardware/docs/html/819-3722-15/index.html#21924 SATA hot plug is supported only for the Windows XP Operating System (OS). Daniel ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss