Re: tmpfs runs out of space on 8.2pre-release, zfs related?
miyamoto moesasji miyamoto.31b at gmail.com writes: In setting up tmpfs (so not tmpmfs) on a machine that is using zfs(v15, zfs v4) on 8.2prerelease I run out of space on the tmpfs when copying a file of ~4.6 GB file from the zfs-filesystem to the memory disk. This machine has 8GB of memory backed by swap on the harddisk, so I expected the file to copy to memory without problems. this is in fact worse than I first thought. After leaving the machine running overnight the tmpfs is reduced to a size of 4K, which shows that tmpfs is in fact completely unusable for me. See the output of df: --- h...@pulsarx4:~/ df -hi /tmp FilesystemSizeUsed Avail Capacity iused ifree %iused Mounted on tmpfs 4.0K4.0K 0B 100% 18 0 100% /tmp --- Relevant zfs-stats info: --- System Memory Statistics: Physical Memory:8161.74M Kernel Memory: 4117.40M DATA: 99.29% 4088.07M TEXT: 0.71% 29.33M ARC Size: Current Size (arcsize): 63.58% 4370.60M Target Size (Adaptive, c): 100.00% 6874.44M Min Size (Hard Limit, c_min): 12.50% 859.31M Max Size (High Water, c_max): ~8:16874.44M --- I'm not sure what triggered this further reduction in size; but the above 4K size is probably important to show how dramatic this goes wrong. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: New ZFSv28 patchset for 8-STABLE
On 01/02/2011 05:06 AM, J. Hellenthal wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 01/01/2011 13:18, Attila Nagy wrote: On 12/16/2010 01:44 PM, Martin Matuska wrote: Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz I've used this: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101223-nopython.patch.xz on a server with amd64, 8 G RAM, acting as a file server on ftp/http/rsync, the content being read only mounted with nullfs in jails, and the daemons use sendfile (ftp and http). The effects can be seen here: http://people.fsn.hu/~bra/freebsd/20110101-zfsv28-fbsd/ the exact moment of the switch can be seen on zfs_mem-week.png, where the L2 ARC has been discarded. What I see: - increased CPU load - decreased L2 ARC hit rate, decreased SSD (ad[46]), therefore increased hard disk load (IOPS graph) Maybe I could accept the higher system load as normal, because there were a lot of things changed between v15 and v28 (but I was hoping if I use the same feature set, it will require less CPU), but dropping the L2ARC hit rate so radically seems to be a major issue somewhere. As you can see from the memory stats, I have enough kernel memory to hold the L2 headers, so the L2 devices got filled up to their maximum capacity. Any ideas on what could cause these? I haven't upgraded the pool version and nothing was changed in the pool or in the file system. Running arc_summary.pl[1] -p4 should print a summary about your l2arc and you should also notice in that section that there is a high number of SPA Mismatch mine usually grew to around 172k before I would notice a crash and I could reliably trigger this while in scrub. What ever is causing this needs desperate attention! I emailed mm@ privately off-list when I noticed this going on but have not received any feedback as of yet. It's at zero currently (2 days of uptime): kstat.zfs.misc.arcstats.l2_write_spa_mismatch: 0 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Specifying root mount options on diskless boot.
--2iBwrppp/7QCDedR Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable [I'm not sure if -stable is the best list for this but anyway...] I'm trying to convert an old laptop running FreeBSD 8.0 into a diskless client (since its internal HDD is growing bad spots faster than I can repair them). I have it pxebooting nicely and running with an NFS root but it then reports locking problems: devd, syslogd, moused (and maybe others) lock their PID file to protect against multiple instances. Unfortunately, these daemons all start before statd/lockd and so the locking fails and reports operation not supported. It's not practical to reorder the startup sequence to make lockd start early enough (I've tried). Since the filesystem is reserved for this client, there's no real need to forward lock requests across the wire and so specifying nolockd would be another solution. Looking through sys/nfsclient/bootp_subr.c, DHCP option 130 should allow NFS mount options to be specified (though it's not clear that the relevant code path is actually followed because I don't see the associated printf()s anywhere on the console. After getting isc-dhcpd to forward this option (made more difficult because its documentation is incorrect), it still doesn't work. Understanding all this isn't helped by kenv(8) reporting three different sets of root filesystem options: boot.nfsroot.path=3D/tank/m3 boot.nfsroot.server=3D192.168.123.200 dhcp.option-130=3Dnolockd dhcp.root-path=3D192.168.123.200:/tank/m3 vfs.root.mountfrom=3Dnfs:server:/tank/m3 vfs.root.mountfrom.options=3Drw,tcp,nolockd And the console also reports conflicting root definitions: Trying to mount root from nfs:server:/tank/m3 NFS ROOT: 192.168.123.200:/tank/m3 Working through all these: boot.nfsroot.* appears to be initialised by sys/boot/i386/libi386/pxe.c but, whilst nfsclient/nfs_diskless.c can parse boot.nfsroot.options, there's no code to initialise that kenv name in pxe.c dhcp.* appears to be initialised by lib/libstand/bootp.c - which does include code to populate boot.nfsroot.options (using vendor specific DHCP option 20) but this code is not compiled in. Further studying of bootp.c shows that it's possible to initialise arbitrary kenv's using DHCP options 246-254 - but the DHCPDISCOVER packets do not request these options so they don't work without special DHCP server configuration (to forward options that aren't requested). vfs.root.* is parsed out of /etc/fstab but, other than being reported in the console message above, it doesn't appear to be used in this environment (it looks like the root entry can be commented out of /etc/fstab without problem). My final solution was to specify 'boot.nfsroot.options=3Dnolockd' in loader.conf - and this seems to actually work. It seems rather unfortunate that FreeBSD has code to allow NFS root mount options to be specified via DHCP (admittedly in several incompatible ways) but none actually work. A quick look at -current suggests that the situation there remains equally broken. Has anyone else tried to use any of this? And would anyone be interested in trying to make it actually work? Hi Peter, i have beed doing diskless booting for a long time, and am very pleased (though 8.2-prerelease is causing some problems :-). In my case /var is mfs, or ufs/zfs, and have no lockd problems. here is what you need to do: either change in libstand/bootp.c: #define DHCP_ENV DHCP_ENV_NO_VENDOR to #define DHCP_ENVDHCP_ENV_FREEBSD or pick my version from: ftp://ftp.cs.huji.ac.il/users/danny/freebsd/diskless-boot/ and compile a new pxeboot. this new pxeboot will allow you to pass via dhcp some key options. next, take a look at ftp://ftp.cs.huji.ac.il/users/danny/freebsd/diskless-boot/rc.initdiskless make sure that your exported root has /.etc If you'r /var is also nfs mounted, maybe unionfs might help too. just writing quickly so you won't feel discouraged, and that diskless actually works. danny ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: slow ZFS on FreeBSD 8.1
On 12/31/2010 6:47 PM, Jeremy Chadwick wrote: On Sat, Jan 01, 2011 at 10:33:43AM +1100, Peter Jeremy wrote: Based on my experiences at home, I converted my desktop at work to pure ZFS. The only issues I've run into have been programs that extensively use mmap(2) - which is a known issue with ZFS. Is your ZFS root filesystem associated with a pool that's mirrored or using raidzX? What about mismatched /boot content (ZFS vs. UFS)? What about booting into single-user mode? http://wiki.freebsd.org/ZFSOnRoot indirectly hints at these problems but doesn't outright admit them (yet should), so I'm curious to know how people have solved them. Remembering manual one-offs for a system configured this way is not acceptable (read: highly prone to error/mistake). Is it worth the risk? Most administrators don't have the tolerance for stuff like that in the middle of a system upgrade or what not; they should be able to follow exactly what's in the handbook, to a tee. There's a link to www.dan.me.uk at the bottom of the above Wiki page that outlines the madness that's required to configure the setup, all of which has to be done by hand. I don't know many administrators who are going to tolerate this when deploying numerous machines, especially when compounded by the complexities mentioned above. This basically outlines the reason why I do not use ZFS on root. -- Dan Langille - http://langille.org/ ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: slow ZFS on FreeBSD 8.1
On Fri, 31 Dec 2010 15:47:47 -0800 Jeremy Chadwick free...@jdc.parodius.com wrote: There's a link to www.dan.me.uk at the bottom of the above Wiki page that outlines the madness that's required to configure the setup, all of which has to be done by hand. I don't know many administrators who are going to tolerate this when deploying numerous machines, especially when compounded by the complexities mentioned above. All of that page could be summarized as: mkdir /usb mount /dev/da4 /usb /usb/install_zfs.sh Where da4 is a USB drive and install_zfs.sh is essentially the commands with some changes needed to support different disks. I'd imagine that administrators wouldn't use sysinstall in interactive mode when deploying to many machines. -- Bruce Cran ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: ZFS - moving from a zraid1 to zraid2 pool with 1.5tb disks
On 1/1/11 6:28 PM, Jean-Yves Avenard wrote: On 2 January 2011 02:11, Damien Fleuriot m...@my.gd wrote: I remember getting rather average performance on v14 but Jean-Yves reported good performance boosts from upgrading to v15. that was v28 :) saw no major difference between v14 and v15. JY Oopsie :) Seeing I for one will have no backups, I think I won't be using v28 on this box, and stick with v15 instead. Are there any views regarding the best implementation for a system ? I currently have a ZFS only system but I'm planning on moving it to UFS, with ZFS used only for mass storage. I understand ZFS root is much trickier, and my main fear is that if I somehow break ZFS (by upgrading to v28 for example) I won't be able to boot anymore, thus no repair process... ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: ZFS - moving from a zraid1 to zraid2 pool with 1.5tb disks
On Sun, 02 Jan 2011 15:31:49 +0100, Damien Fleuriot m...@my.gd wrote: On 1/1/11 6:28 PM, Jean-Yves Avenard wrote: On 2 January 2011 02:11, Damien Fleuriot m...@my.gd wrote: I remember getting rather average performance on v14 but Jean-Yves reported good performance boosts from upgrading to v15. that was v28 :) saw no major difference between v14 and v15. JY Oopsie :) Seeing I for one will have no backups, I think I won't be using v28 on this box, and stick with v15 instead. Are there any views regarding the best implementation for a system ? I currently have a ZFS only system but I'm planning on moving it to UFS, with ZFS used only for mass storage. I understand ZFS root is much trickier, and my main fear is that if I somehow break ZFS (by upgrading to v28 for example) I won't be able to boot anymore, thus no repair process... You can repair by booting from USB of CD in a lot of cases. Ronald. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: ZFS - moving from a zraid1 to zraid2 pool with 1.5tb disks
On 2010-Dec-30 12:40:00 +0100, Damien Fleuriot m...@my.gd wrote: What are the steps for properly removing my drives from the zraid1 pool and inserting them in the zraid2 pool ? I've documented my experiences in migrating from a 3-way RAIDZ1 to a 6-way RAIDZ2 at http://bugs.au.freebsd.org/dokuwiki/doku.php/zfsraid Note that, even for a home system, backups are worthwhile. In my case, I backup onto a 2TB disk in an eSATA enclosure. That's currently (just) adequate but I'll soon need to identify data that I can leave off that backup. -- Peter Jeremy pgpOSt5NCO7Do.pgp Description: PGP signature
Re: tmpfs runs out of space on 8.2pre-release, zfs related?
On Sun, 02 Jan 2011 09:41:04 +0100, miyamoto moesasji miyamoto@gmail.com wrote: miyamoto moesasji miyamoto.31b at gmail.com writes: In setting up tmpfs (so not tmpmfs) on a machine that is using zfs(v15, zfs v4) on 8.2prerelease I run out of space on the tmpfs when copying a file of ~4.6 GB file from the zfs-filesystem to the memory disk. This machine has 8GB of memory backed by swap on the harddisk, so I expected the file to copy to memory without problems. this is in fact worse than I first thought. After leaving the machine running overnight the tmpfs is reduced to a size of 4K, which shows that tmpfs is in fact completely unusable for me. See the output of df: --- h...@pulsarx4:~/ df -hi /tmp FilesystemSizeUsed Avail Capacity iused ifree %iused Mounted on tmpfs 4.0K4.0K 0B 100% 18 0 100% /tmp --- Relevant zfs-stats info: --- System Memory Statistics: Physical Memory:8161.74M Kernel Memory: 4117.40M DATA: 99.29% 4088.07M TEXT: 0.71% 29.33M ARC Size: Current Size (arcsize): 63.58% 4370.60M Target Size (Adaptive, c): 100.00% 6874.44M Min Size (Hard Limit, c_min): 12.50% 859.31M Max Size (High Water, c_max): ~8:16874.44M --- I'm not sure what triggered this further reduction in size; but the above 4K size is probably important to show how dramatic this goes wrong. Is it possible that some program is filling a file (on your tmpfs) which is deleted? You will not see it with df or ls, but it still takes space on the fs, until the application closes the file. Ronald. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: tmpfs runs out of space on 8.2pre-release, zfs related?
On Sun, Jan 2, 2011 at 3:26 PM, Ronald Klop ronald-freeb...@klop.yi.org wrote: On Sun, 02 Jan 2011 09:41:04 +0100, miyamoto moesasji miyamoto@gmail.com wrote: miyamoto moesasji miyamoto.31b at gmail.com writes: In setting up tmpfs (so not tmpmfs) on a machine that is using zfs(v15, zfs v4) on 8.2prerelease I run out of space on the tmpfs when copying a file of ~4.6 GB file from the zfs-filesystem to the memory disk. This machine has 8GB of memory backed by swap on the harddisk, so I expected the file to copy to memory without problems. this is in fact worse than I first thought. After leaving the machine running overnight the tmpfs is reduced to a size of 4K, which shows that tmpfs is in fact completely unusable for me. See the output of df: --- h...@pulsarx4:~/ df -hi /tmp Filesystem Size Used Avail Capacity iused ifree %iused Mounted on tmpfs 4.0K 4.0K 0B 100% 18 0 100% /tmp --- Relevant zfs-stats info: --- System Memory Statistics: Physical Memory: 8161.74M Kernel Memory: 4117.40M DATA: 99.29% 4088.07M TEXT: 0.71% 29.33M ARC Size: Current Size (arcsize): 63.58% 4370.60M Target Size (Adaptive, c): 100.00% 6874.44M Min Size (Hard Limit, c_min): 12.50% 859.31M Max Size (High Water, c_max): ~8:1 6874.44M --- I'm not sure what triggered this further reduction in size; but the above 4K size is probably important to show how dramatic this goes wrong. Is it possible that some program is filling a file (on your tmpfs) which is deleted? You will not see it with df or ls, but it still takes space on the fs, until the application closes the file. Ronald. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org I'm pretty sure that this is not the case as the problem shows up immediately upon a clean reboot (and /tmp is normally pretty empty with just 22k in use at the moment with the older tmpmfs (determined with du -xh /tmp as root) ) Note that the key problem was given in the first post. There I posted that the tmpfs has 8.2GB available immediately after a reboot, yet it is impossible to copy a 4.6GB file to tmpfs from a zfs-drive even though the machine has 8GB of memory backed with swap. My feeling is that somehow the zfs file cache, which is also in memory is causing this, which is consistent with the solaris bugreport I linked to, see: http://bugs.opensolaris.org/bugdatabase/view_bug.do;jsessionid=e4ae9c32983000ef651e38edbba1?bug_id=6804661 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: slow ZFS on FreeBSD 8.1
- Original Message - From: Jeremy Chadwick free...@jdc.parodius.com Based on my experiences at home, I converted my desktop at work to pure ZFS. The only issues I've run into have been programs that extensively use mmap(2) - which is a known issue with ZFS. Is your ZFS root filesystem associated with a pool that's mirrored or using raidzX? What about mismatched /boot content (ZFS vs. UFS)? What about booting into single-user mode? http://wiki.freebsd.org/ZFSOnRoot indirectly hints at these problems but doesn't outright admit them (yet should), so I'm curious to know how people have solved them. Remembering manual one-offs for a system configured this way is not acceptable (read: highly prone to error/mistake). Is it worth the risk? Most administrators don't have the tolerance for stuff like that in the middle of a system upgrade or what not; they should be able to follow exactly what's in the handbook, to a tee. There's a link to www.dan.me.uk at the bottom of the above Wiki page that outlines the madness that's required to configure the setup, all of which has to be done by hand. I don't know many administrators who are going to tolerate this when deploying numerous machines, especially when compounded by the complexities mentioned above. With regards installing machines with a root zfs we now use mfsBSD which makes the process as simple as pie, so for those that haven't used it give it a wirl:- http://mfsbsd.vx.sk/ The mmap(2) and sendfile(2) complexities will bite an junior or mid-level SA in the butt too -- they won't know why software starts failing or behaving oddly (FreeBSD ftpd is a good example). It just so happens that Apache, out-of-the-box, comes with mmap and sendfile use disabled. This is the same with nginx which is rapidly taking over from apache due to its ability to scale much much better than apache does. Proper mmap and sendfile integration are the only major issue we have with moving all our machines to ZFS thanks to great work by everyone. I really hope sendfile support in particular is fixed in the near future but as I understanding it, that's not going to be simple at all :( Regards Steve This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Sense fetching [Was: cdrtools /devel ...]
If FreeBSD writes any warning for SCSI commands send by cdrtools, then this is definitely a kernel bug that needs to be fixed. SCSI is a protocol that lives from aparently failing SCSI commands. These warnings are related to high level interaction between cdrecord and the drive. Only cdrecord is able to decide whether a SCSI command that apparently failed is worth for being reported or not. It therefore is a bug if the kernel prints related messages. Jörg -- EMail:jo...@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin j...@cs.tu-berlin.de(uni) joerg.schill...@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/ URL: http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: umass: AutoSense failed
On Tue, Dec 14, 2010 at 01:19:29PM -0700, Ross Alexander wrote: On Sat, 11 Dec 2010, Harad Weiss wrote Message: 7 Date: Fri, 10 Dec 2010 22:59:17 +0100 From: Harald Weis ha...@free.fr Subject: Re: umass: AutoSense failed To: freebsd-stable@freebsd.org Cc: Jeremy Chadwick free...@jdc.parodius.com Message-ID: 20101210215917.ga2...@pollux.local.net Content-Type: text/plain; charset=us-ascii [big snip] The AutoSense failure occurs on two desktop computers, whether motherboard or not, and on three laptops. As I said, all PC's run 8.1-RELEASE. What I do not know is whether the Sony DSC (Digital Still Camera) has worked on 8.0-RELEASE. I didn't use it for some time, so I'm afraid I never tried it on 8.x. But I am sure it worked on all earlier releases. In my original post I have forgotten to mention that I never had any problems with all sorts of thumbdrives and USB disks. Fortunately, I can read the 256MB CF memory of the DSC on an Ubuntu laptop. Could it possibly be a DSC hardware problem which is ignored by Ubuntu, but not by FreeBSD? My Sony DSC camera did work under 8.0-RELEASE. It's a model DSC-W35. It's seeing the umass problem now, and has since 8.1-PRERELEASE afair. This fault is the same on i386 or amd64, intel or amd chipsets, USB 1.0 or 2.0, through hubs or straight off the m/b, etc.. I'm keeping a backup box (intel atom / i386) running 7.3-STABLE around specifically to work around the problem. I'd have said something eralier, but I've gotten used to small nits poppping up and then disappearing; this one is dragging along :( I haven't looked at the List for a while, so I only saw your message yesterday. In the mean time, I've re-learned to work with bootable memory sticks. I've had no luck yet with ReWritable CDs which I need for all boxes but one. Perhaps a blanking problem? Here is the summary of my trials in Fixit mode for my Sony DSC-P10 which is also (at last) the answer to Jeremy's message: 7.2-RELEASE-i386-livefs: OK 8.0-RELEASE-i386-memstick: OK FreeBSD-8.1-STABLE-201009-i386-memstick: AutoSense failed FreeBSD-8.2-PRERELEASE-201012-i386-memstick: booting OK, but No USB devices found! message extract CAM status: SCSI Status Error SCSI sense: UNIT ATTENTION asc: 28,0 (Not ready to ready change, medium may have changed) da0: 1915MB (3922944 512 byte sectors: 255H 63S/T 244C) GEOM: da0: media size does not match label. /message extract FreeBSD-9.0-CURRENT-201009-i386-memstick: AutoSense failed Is it not a case for send-pr(1) ? Best regards, Harald ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: New ZFSv28 patchset for 8-STABLE
On 01/02/2011 03:45, Attila Nagy wrote: On 01/02/2011 05:06 AM, J. Hellenthal wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 01/01/2011 13:18, Attila Nagy wrote: On 12/16/2010 01:44 PM, Martin Matuska wrote: Link to the patch: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101215.patch.xz I've used this: http://people.freebsd.org/~mm/patches/zfs/v28/stable-8-zfsv28-20101223-nopython.patch.xz on a server with amd64, 8 G RAM, acting as a file server on ftp/http/rsync, the content being read only mounted with nullfs in jails, and the daemons use sendfile (ftp and http). The effects can be seen here: http://people.fsn.hu/~bra/freebsd/20110101-zfsv28-fbsd/ the exact moment of the switch can be seen on zfs_mem-week.png, where the L2 ARC has been discarded. What I see: - increased CPU load - decreased L2 ARC hit rate, decreased SSD (ad[46]), therefore increased hard disk load (IOPS graph) Maybe I could accept the higher system load as normal, because there were a lot of things changed between v15 and v28 (but I was hoping if I use the same feature set, it will require less CPU), but dropping the L2ARC hit rate so radically seems to be a major issue somewhere. As you can see from the memory stats, I have enough kernel memory to hold the L2 headers, so the L2 devices got filled up to their maximum capacity. Any ideas on what could cause these? I haven't upgraded the pool version and nothing was changed in the pool or in the file system. Running arc_summary.pl[1] -p4 should print a summary about your l2arc and you should also notice in that section that there is a high number of SPA Mismatch mine usually grew to around 172k before I would notice a crash and I could reliably trigger this while in scrub. What ever is causing this needs desperate attention! I emailed mm@ privately off-list when I noticed this going on but have not received any feedback as of yet. It's at zero currently (2 days of uptime): kstat.zfs.misc.arcstats.l2_write_spa_mismatch: 0 Right but do you have a 'cache' 'l2arc' vdev attached to any pool in the system ? This suggests to me that you do not at this time. If not can you attach a cache vdev and run a scrub on it and monitor the value of that MIB ? -- Regards, jhell,v JJH48-ARIN ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org