Date: Tue, 31 Jul 2012 15:27:57 +0200 (CEST) From: Kasper Sacharias Eenberg <kasper@cabo.dk> To: freebsd-fs@freebsd.org Subject: Kernel panick 'vm_map_entry_create' when copying to ZFS volume shared over NFS. Message-ID: <29509988.197565.1343741277744.JavaMail.root@zmbox01> In-Reply-To: <20759432.197480.1343738543702.JavaMail.root@zmbox01>
next in thread | previous in thread | raw e-mail | index | archive | help
Problem: On copying files from vmware sphere client to a FreeBSD NFS shared ZFS volume, I get the unwanted error 'vm_map_entry_create'. A screenshot of the KVM viewer displaying the error can be found here: http://i.imgur.com/T1AnZ.jpg I can't really go into detail about how the copying is done, but it shouldn't matter. The problem is reproducible, it's happened 3 times so far. I'm about to test with and even lower vfs.zfs.arc_max, since it happens at about 14->15 GB arc_size, I've also removed the log-ssd, which is the device reset warning in the screenshot. I encountered the same error. Having to tune arc_max manually to get a working system seems quite silly. (If it works..) The harddrives are 7 OCZ Vertex4 SSD's, in raidz. The log drive is of the same type. Sorry about the length of the mail, I've added all I though might help. The system logs contain nothing. Does anyone have any input? Software: FreeBSD 9.0-RELEASE with added Intel ixgbe-2.4.4 driver module loaded (and used). ZFS volume information NAME PROPERTY VALUE pool-ssd/data001 type filesystem pool-ssd/data001 creation Tue Jun 19 10:04 2012 pool-ssd/data001 used 99.9G pool-ssd/data001 available 2.63T pool-ssd/data001 referenced 35.8G pool-ssd/data001 compressratio 1.54x pool-ssd/data001 mounted yes pool-ssd/data001 quota none pool-ssd/data001 reservation none pool-ssd/data001 recordsize 8K pool-ssd/data001 mountpoint /pool-ssd/data001 pool-ssd/data001 sharenfs -mapall=root -network <IP> -mask 255.255.255.0 pool-ssd/data001 checksum on pool-ssd/data001 compression lzjb pool-ssd/data001 atime off pool-ssd/data001 devices on pool-ssd/data001 exec on pool-ssd/data001 setuid on pool-ssd/data001 readonly off pool-ssd/data001 jailed off pool-ssd/data001 snapdir hidden pool-ssd/data001 aclmode discard pool-ssd/data001 aclinherit restricted pool-ssd/data001 canmount on pool-ssd/data001 xattr off pool-ssd/data001 copies 1 pool-ssd/data001 version 5 pool-ssd/data001 utf8only off pool-ssd/data001 normalization none pool-ssd/data001 casesensitivity sensitive pool-ssd/data001 vscan off pool-ssd/data001 nbmand off pool-ssd/data001 sharesmb off pool-ssd/data001 refquota none pool-ssd/data001 refreservation none pool-ssd/data001 primarycache all pool-ssd/data001 secondarycache all pool-ssd/data001 usedbysnapshots 64.1G pool-ssd/data001 usedbydataset 35.8G pool-ssd/data001 usedbychildren 0 pool-ssd/data001 usedbyrefreservation 0 pool-ssd/data001 logbias latency pool-ssd/data001 dedup off pool-ssd/data001 mlslabel pool-ssd/data001 sync standard pool-ssd/data001 refcompressratio 1.25x A second before the crash: ------------------------------------------------------------------------ ZFS Subsystem ReportTue Jul 31 13:58:34 2012 ------------------------------------------------------------------------ System Memory: 0.13%30.15MiB Active,0.02%5.82MiB Inact 67.53%15.69GiB Wired,0.00%520.00KiB Cache 32.31%7.51GiB Free,0.00%404.00KiB Gap Real Installed:24.00GiB Real Available:99.88%23.97GiB Real Managed:96.94%23.24GiB Logical Total:24.00GiB Logical Used:68.69%16.49GiB Logical Free:31.31%7.51GiB Kernel Memory:13.26GiB Data:99.85%13.24GiB Text:0.15%20.09MiB Kernel Memory Map:20.24GiB Size:65.31%13.22GiB Free:34.69%7.02GiB ------------------------------------------------------------------------ ARC Summary: (HEALTHY) Memory Throttle Count:0 ARC Misc: Deleted:15 Recycle Misses:0 Mutex Misses:0 Evict Skips:0 ARC Size:62.93%13.99GiB Target Size: (Adaptive)100.00%22.24GiB Min Size (Hard Limit):12.50%2.78GiB Max Size (High Water):8:122.24GiB ARC Size Breakdown: Recently Used Cache Size:58.97%13.11GiB Frequently Used Cache Size:41.03%9.12GiB ARC Hash Breakdown: Elements Max:1.70m Elements Current:100.00%1.70m Collisions:1.20m Chain Max:14 Chains:439.54k ------------------------------------------------------------------------ A second before the crash: ------------------------------------------------------------------------ ZFS Subsystem ReportTue Jul 31 14:40:26 2012 ------------------------------------------------------------------------ System Memory: 0.12%28.21MiB Active,0.02%5.27MiB Inact 69.62%16.18GiB Wired,0.00%40.00KiB Cache 30.24%7.03GiB Free,0.00%444.00KiB Gap Real Installed:24.00GiB Real Available:99.88%23.97GiB Real Managed:96.94%23.24GiB Logical Total:24.00GiB Logical Used:70.70%16.97GiB Logical Free:29.30%7.03GiB Kernel Memory:13.70GiB Data:99.86%13.69GiB Text:0.14%20.09MiB Kernel Memory Map:20.19GiB Size:67.65%13.66GiB Free:32.35%6.53GiB ------------------------------------------------------------------------ ARC Summary: (HEALTHY) Memory Throttle Count:0 ARC Misc: Deleted:15 Recycle Misses:0 Mutex Misses:0 Evict Skips:0 ARC Size:96.34%14.45GiB Target Size: (Adaptive)100.00%15.00GiB Min Size (Hard Limit):12.50%1.88GiB Max Size (High Water):8:115.00GiB ARC Size Breakdown: Recently Used Cache Size:89.54%13.43GiB Frequently Used Cache Size:10.46%1.57GiB ARC Hash Breakdown: Elements Max:1.74m Elements Current:100.00%1.74m Collisions:1.24m Chain Max:14 Chains:443.12k ------------------------------------------------------------------------ /boot/loader.conf --------------------- ahci_load="yes" zfs_load="YES" if_vlan_load="YES" vfs.root.mountfrom="zfs:zroot" ixgbe_load="YES" boot_multicons="YES" boot_serial="YES" comconsole_speed="115200" console="comconsole,vidconsole" hw.igb.rxd="4096" hw.igb.txd="4096" kern.ipc.nmbclusters="262144" kern.ipc.nmbjumbop="262144" # Disable ZFS prefetching # http://southbrain.com/south/2008/04/the-nightmare-comes-slowly-zfs.html # Increases overall speed of ZFS, but when disk flushing/writes occur, # system is less responsive (due to extreme disk I/O). # NOTE: 8.0-RC1 disables this by default on systems <= 4GB RAM anyway vfs.zfs.prefetch_disable="1" # Disable UMA (uma(9)) for ZFS; amd64 was moved to exclusively use UMA # on 2010/05/24. # http://lists.freebsd.org/pipermail/freebsd-stable/2010-June/057162.html #vfs.zfs.zio.use_uma="0" # Decrease ZFS txg timeout value from 30 (default) to 5 seconds. This # should increase throughput and decrease the "bursty" stalls that # happen during immense I/O with ZFS. # http://lists.freebsd.org/pipermail/freebsd-fs/2009-December/007343.html # http://lists.freebsd.org/pipermail/freebsd-fs/2009-December/007355.html vfs.zfs.txg.timeout="5" # Below was added to test whether limiting arc memory worked. vfs.zfs.arc_max="15G" --------------------- Kind regards / Med venlig hilsen Kasper Sacharias Eenberg Cabo A/S Klosterport 4a, 4. sal 8000 Aarhus C
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?29509988.197565.1343741277744.JavaMail.root>