Patch-ID# 116669-23 NOTE: *********************************************************************** READ THE TERMS OF THE AGREEMENT ("AGREEMENT") IN THE LEGAL_LICENSE.TXT FILE CAREFULLY BEFORE USING THIS SOFTWARE. BY USING THE SOFTWARE, YOU AGREE TO THE TERMS OF THIS AGREEMENT. IF YOU DO NOT AGREE TO ALL OF THE TERMS, PROMPTLY DESTROY THE UNUSED SOFTWARE. *********************************************************************** Keywords: md rcm libsvmjni dr metadevice svm libmeta rcm metadb corruption Synopsis: SunOS 5.9: md Patch Date: Jan/16/2007 Install Requirements: Reconfigure immediately after patch is installed Install in Single User Mode Solaris Release: 9 SunOS Release: 5.9 Unbundled Product: Unbundled Release: Xref: This patch available for x86 as patch 113994 and 118559 Topic: SunOS 5.9: md Patch Relevant Architectures: sparc BugId's fixed with this patch: 4380147 4634775 4695547 4734624 4750899 4790035 4813460 4822189 4826485 4842627 4886687 4892666 4915431 4927518 4951103 4964361 4981298 5018506 5026655 5047567 5048347 5048915 5061410 5061599 5063525 5067898 5068997 5070446 5070496 5070511 5071059 5071901 5072849 5073220 5073727 5080192 5080292 5081041 5081477 5085032 5087369 5088893 5089083 5089909 5093712 5097853 5102518 5109446 5109456 5110153 6176220 6178538 6178560 6182797 6184792 6189856 6192304 6209065 6217370 6220234 6220478 6224745 6226297 6242768 6243675 6244431 6290815 6304858 6324927 6341629 6345556 6370127 6448006 6448544 Changes incorporated in this version: 6448544 Patches accumulated and obsoleted by this patch: 115559-01 119432-01 Patches which conflict with this patch: Patches required with this patch: 113073-14 (or greater) Obsoleted by: Files included with this patch: /kernel/drv/md /kernel/drv/sparcv9/md /kernel/misc/md_mirror /kernel/misc/md_raid /kernel/misc/md_sp /kernel/misc/md_stripe /kernel/misc/md_trans /kernel/misc/sparcv9/md_mirror /kernel/misc/sparcv9/md_raid /kernel/misc/sparcv9/md_sp /kernel/misc/sparcv9/md_stripe /kernel/misc/sparcv9/md_trans /sbin/metadb /sbin/metadevadm /sbin/metainit /sbin/metarecover /sbin/metastat /usr/include/mdiox.h /usr/include/meta.h /usr/include/meta_basic.h /usr/include/metad.h /usr/include/sys/lvm/md_basic.h /usr/include/sys/lvm/md_mddb.h /usr/include/sys/lvm/md_mdiox.h /usr/include/sys/lvm/md_mirror.h /usr/include/sys/lvm/mdio.h /usr/include/sys/lvm/mdmn_commd.h /usr/include/sys/lvm/mdvar.h /usr/lib/libmeta.so.1 /usr/lib/rcm/modules/SUNW_svm_rcm.so /usr/sbin/metaclear /usr/sbin/metaimport /usr/sbin/metainit /usr/sbin/metaset /usr/sbin/metasync /usr/sbin/metattach /usr/sbin/rpc.mdcommd /usr/sbin/rpc.metad Problem Description: 6448544 diskset minor name namespace records are not correct (from 116669-22) 6448006 4 node system paniced after adding some more disks to an existing metaset (from 116669-21) 4951103 assertion error on reboot after disabling MPxIO 6226297 don't retry failfast IO unless device is Last Erred (from 116669-20) 5018506 mbi_mddb_mb Null pointer referenced in getphysblk panics the system 6304858 S10 cluster lost metadb (from 116669-19) 6243675 creating a soft partition with -o value of less than the interlace size will fail (from 116669-18) 6341629 metaset -s diskset -t dumps core with devid-enabled DID driver on Sun Cluster node 6370127 SVM mediators are not updated in all places they should be (from 116669-17) 6242768 metaset cannot take disk set after metad client create : RPC: Timed out (from 116669-16) 6324927 multi-owner diskset: if any mediator fails to respond, nodes not able to join the set 6345556 get_db_rec needs to use correct ioctl for MN sets (from 116669-15) 5081041 refresh of inetd can cause SVM RPC timeouts 5089083 cfgadm: Library error: RCM request delete capacity failed for SB9 6290815 SVM exited with error 1 in step cmmstep4, nodes panic 6192304 metainit of sp corrupts the dsk label of the 1st dsk in stripe if vtoc and stripe > 1TB 6220478 disks with different devts cannot be added to multi-node disksets (from 116669-14) 5080192 SVM gets very confused with multipath disks when not using MPxIO 6184792 failure in ddi call in md_resolve_bydevid could cause return of incorrect data (from 116669-13) 4734624 metastat dumps core when aborting the TSmd test suite (from 116669-12) 5070446 unable to take a diskset in "boot -x" mode on Sun Cluster 3 with s9u7 (from 116669-11) 4695547 SVM soft partition alignment can cause additional physical disk IOs 5026655 metainit -p -e fails for disk with EFI label and metadb replica 4826485 mdiox.x contains C syntax error 4750899 when adding an EFI disk to a diskset, it always repartitions 4842627 cut and paste duplicate strings in meta_error.c 6189856 segmentation fault when creating a soft partition of 0 blocks 4634775 PSARC/2002/509 Maxsize flag for metainit of Soft Partitions (from 116669-10) 6220234 failfast flag is never set on writes 6244431 SVM - metadbs are not found after reboot when FUJITSU MAT3073N or ST373207LSUN72G disks are used (from 116669-09) 4927518 SVM could be more intelligent when components are being unconfigured 4892666 minor coding errors in SVM rcm module 6224745 unconfiguring raid component puts it into maintenance 6217370 raid 1+0 hangs in fsck when offsetting slices are errored 5047567 enable device id information for block copied SVM disks 4964361 metaimport should handle block-copied disks 5063525 metaset -t fail for traditional set with more than one host (from 116669-08) 5110153 metattach - does not update the md.cf file when used to grow soft partitions (from 116669-07) 5073220 problem when adding a Sun Cluster EFI DID disk to a metaset (from 116669-06) 4790035 md causes blocking chain of doom 5061599 ucmmd dumps core and the node is panicked by failfast 5067898 both submirrors stuck in "Needs maintenance" following node failure 5070496 x86 Cluster node panics during high stress tests 5081477 forced panic when metadetach returns different values on different nodes 5087369 additional debug information for rpc.mdcommd 5088893 svmreconfig timeout lead to cluster node panic 5093712 master node panic if metacommand in progress when slave node fails 5102518 panic MD_SP_ERR could not be set on all nodes 5109446 on Multi-Node sets, SVM does not correctly deal with 3-way mirrors when 2 fail 5109456 soft partition watermarks flagged as incorrect during mirror_stress test 6176220 SVM may hang when waiting for a resync thread to terminate when starting a new resync 6178538 creating a single MN diskset metadevice using md.tab format creates metadevice on only 1 node 6178560 MN diskset metaclust step1 returns 205 when multiple starting nodes run step1 simultaneously 6182797 MN diskset data loss can occur if node panics during disk deletion (from 116669-05) 5089909 all meta* commands will core dump (from 116669-04) 4886687 boot panic if <=50% MDBs and root mirror hot swaps to unavailable hot spare 4981298 domain panic during concurrent SVM and DR operations 5071059 SVM panic when performing metareplace 5068997 panic in md_layered_close due to NODEV being passed in 5073727 assertion failed: this_dev != NODEV64 (from 116669-03) 4915431 SVM should stop talking trash 5097853 SVM s9 patches missing SUNW_svm_rcm.so (from 116669-02) 5048347 panic when concurrently running cluster reconfig cycle and metadetach 5048915 console message preceding forced panic from rpc.mdcommd missing 5061410 mirror stuck in "Needs maintenance" when it should be resyncing 5070511 panic during the Cycle2 automated test 2.1.1 5071901 metaoffline for a mirror with a soft part in ABR mode will cause data corruption 5072849 race between metaclear and hotspares causes the machine to panic 5080292 using metattach to grow a mirror does not work on a multi-owner disk set 5085032 bool_t return type missing from xdr_ushort_t, xdr_dev_t in meta_basic.x (from 116669-01) 4380147 metaset takeover under mediator takes too long... (from 115559-01) 4813460 it took ~30-40 mins to DR an IO board with a freshchoice+network traffic+ipmp 4822189 unconfiguring a child of a metadevice does not yield useful error message (from 119432-01) 6209065 metadb corruption breaks Solaris development machine (AMD64) multiple times Patch Installation Instructions: -------------------------------- For Solaris 2.0-2.6 releases, refer to the Install.info file and/or the README within the patch for instructions on using the generic 'installpatch' and 'backoutpatch' scripts provided with each patch. For Solaris 7-10 releases, refer to the man pages for instructions on using 'patchadd' and 'patchrm' scripts provided with Solaris. Any other special or non-generic installation instructions should be described below as special instructions. The following example installs a patch to a standalone machine: example# patchadd /var/spool/patch/104945-02 The following example removes a patch from a standalone system: example# patchrm 104945-02 For additional examples please see the appropriate man pages. Special Install Instructions: ----------------------------- NOTE 1: Perform patch installation in single user mode. Perform a reconfiguration boot, boot -r, after patch installation. NOTE 2: For the complete fix to 4380147 (metaset takeover under mediator takes too long), please also install the following patch: 113278-07 (or greater) NFS Daemon, rpcmod patch README -- Last modified date: Tuesday, January 16, 2007