|
Patch 2851 : XLV Rollup #5 : [IRIX 6.2]
INDEX
RELATIONS
RELEASE NOTES
1. Patch SG0002851 Release Note
This release note describes patch SG0002851 to IRIX 6.2.
1.1 Supported Hardware Platforms
This patch contains bug fixes for all platforms supported
under Irix 6.2. The software cannot be installed on other
configurations.
1.2 Supported Software Platforms
This patch contains bug fixes for XLV on a system running
IRIX 6.2. The software cannot be installed on other
configurations.
1.3 Bugs Fixed by Patch SG0002851
This patch contains fixes for the following bugs in IRIX
6.2.
o The xlv_mgr(1M) command could core when both the volume
name and the nodename were the same. This could happen
only by removing a volume, naming a new one
identically, and then subsequently reintroducing the
old one. It now prints an error message that the
second one found has been ignored. (bug #537917)
o Incomplete volume elements can cause a kernel panic.
Occurs on mirrored plexes where different portions of
the address range are valid on the different plex
halves. (bug #548381)
o Systems with a large number of disks could not
successfully read the plex license file. This is an
O32 ABI issue where more than 255 file descriptors
cannot be used by fopen(3). (bug #558182)
This patch also "rolls up" patch 2457 that fixes the bugs
list below:
o The XLV label daemon is pinned in memory and allocates
and frees a large structure (2MB). Since it is pinned,
the heap never shrinks, causing xlv_labd(1m) to be a
memory hog. (bug #438539)
o During system startup, if /tmp is a volume and there
are plexed volumes being revived, the mount of the /tmp
volume fails. /tmp cannot mount until all the plexed
volumes are revived. (bug# 471909)
o During a XLV plex revive, a read error on the source
volume element causes the system to panic with a kernel
fault. (bug# 474748)
o XLV doesn't process multiple plex errors correctly when
receiving many disk errors, and can cause a kernel
panic. (bug #528089)
o A root volume which is either concatenated or striped
isn't properly rejected at boot time. (bug# 532078)
o xlv_mgr(1m) doesn't handle volumes with duplicate names
but different hostnames. (bug #533693)
o The death of the xlv_labd(1m) process can cause the
kernel to hang. We work around this by restarting the
deamon as necessary. (bug #534748)
This patch also "rolls up" patch 1613 that fixes the bugs
list below:
o XLV can deadlock the system under high load when it
grabs locks which prevent I/O while running
configuration commands (i.e, xlv_assemble). The
xlv_assemble process may need then to acquire memory,
causing the system to need to do I/O, resulting in
deadlock. (bug #375080)
o Increased the number of opened volume headers file
descriptors the XLV utilities can have. The old limit
was 16.
o Added a create command to xlv_make(1m) for writing
newly defined objects out to affected disks without
exiting the tool. (bug #250074)
o Added three new commands to xlv_mgr(1m). The change
name command renames a xlv object. The change type
command changes the type of an object. The change
ve_start command changes start address of a standalone
volume element. (bug #331525)
o Don't allow xlv_mgr(1m) to change an active plex to
empty using the change online command. In order to
force a revive, the correct procedure is to issue a
change offline, followed by change online.
o Corrected misleading xlv_make(1m) error message
regarding missing block and/or character device node.
(bug #267257)
o Added a -h command line option to xlv_make(1m) so a
nodename can be specified. This option implies the -A
option to not automatically invoke xlv_assemble(1m).
(bug #312367)
o The xlv_shutdown(1m) -n option takes a comma separated
list of volume names so multiple volumes can be shut
down.
o Doing a selective xlv_shutdown (-n option) when there
are any volumes with log subvolumes could cause xlv
label corruption.
o Fixed a problem in xlv_shutdown(1m) which caused it to
fail with out of memory errors on large systems. (bug
#405757)
o Xlv_shutdown(1m) failed to shut down a volume when the
volume did not have any active volume elements.
o In two different windows, simultaneously using
xlv_shutdown(1m) and xlv_mgr(1m) to remove a volume can
cause the a system panic. (bug #382630)
o The -a option to xlv_assemble(1m) was broken in patch
1228. New volumes could not be configured into the
kernel. (bug #389537)
o Changed xlv_assemble to exit with a non-zero exit
status if selective assemble can't assemble all of the
volumes specified on the command line. (bug #422453)
o The xlv_mgr(1m) script all did not correctly generate
the time of day. Sometimes the command caused a core
dump.
o xlv_labd(1m) is automatically started if a plex revive
request is issued and the xlv_labd does not exist.
o Losing a disk drive that is part of a root volume can
sometimes cause the system to hang. (bug #392195)
o A plexed root drive could be corrupted if one plex was
detached/attached, and the system was then cleanly
shutdown and rebooted. (bug #422798)
o Xlv_make(1m) failed to create a standalone plex or
volume element when the inter-mixing disk partitions of
types "xlv" and "xfs". (bug #422766)
o Xlv_mgr(1m) core dumps detaching a NULL volume element
from a standalone plex. (bug #426966)
o Only mark the target disk online after a revive if no
errors occured.
This patch also "rolls up" patch 1228 that fixes the bugs
list below:
o Added a -s option to xlv_assemble(1m) so a list of
logical volumes to assemble can be specified on the
command line. It is no longer necessary to assemble
all logical volumes at once. This option assists
failover procedures for the FailSafe Product. (bug
#311156 & bug #333694).
o XLV device nodes were created even if the volume was
incomplete, possibly leading to the mistaken conclusion
that the volume was usable. (bug #294565)
o Xlv_assemble(1m) won't gratuitously initiate a plex
revive for a triple-plexed volume element when the
volume element was cleanly shutdown. (bug #341023)
o XLV displays the nodename assoicated with a volume.
(bug #290849)
o Xlv_mgr(1m) "detach" command has a new -force option so
a missing plex or volume element can be removed. This
allows broken disks to be easily replaced.
o Xlv_mgr(1m) "show kernel" command displays plex revive
progress. The mapping of the subvolume open_flag value
has been corrected. A volume name can be specified so
that only information pertaining to that kernel volume
is displayed.
o Xlv_make(1m) and xlv_mgr(1m) have a new -c option for
specifying input on the command line instead of the
interactive interface.
o The system can sometimes panic when a subvolume goes
offline. (bug #355237 & bug #357554)
o Xlv_shutdown(1m) complains and fails with a exit code
of 1 when an invalid volume name is specified with the
-n option. (bug #323922)
o Fixes a kernel panic if plexing software is not
installed and an I/O error occurs on a RAID device.
(bug #363121)
o Fail over to an alternate controller on a dual
controller SGI RAID device only works on lun 0. Luns
other than 0 were not properly switched over to the
alternate path. (bug #370878)
1.4 Subsystems Included in Patch SG0002851
This patch release includes these subsystems:
o patchSG0002851.eoe_sw.unix
o patchSG0002851.eoe_sw.xlv
o patchSG0002851.eoe_sw.xlvplex
1.5 Installation Instructions
Because you want to install only the patches for problems
you have encountered, patch software is not installed by
default. After reading the descriptions of the bugs fixed
in this patch (see Section 1.3), determine the patches that
meet your specific needs.
If, after reading Sections 1.1 and 1.2 of these release
notes, you are unsure whether your hardware and software
meet the requirements for installing a particular patch, run
inst. The inst program does not allow you to install
patches that are incompatible with your hardware or
software.
Patch software is installed like any other Silicon Graphics
software product. Follow the instructions in your Software
Installation Administrator's Guide to bring up the miniroot
form of the software installation tools.
Follow these steps to select a patch for installation:
1. At the Inst> prompt, type
install patchSGxxxxxxx
where xxxxxxx is the patch number.
2. Initiate the installation sequence. Type
Inst> go
3. You may find that two patches have been marked as
incompatible. (The installation tools reject an
installation request if an incompatibility is
detected.) If this occurs, you must deselect one of
the patches.
Inst> keep patchSGxxxxxxx
where xxxxxxx is the patch number.
4. After completing the installation process, exit the
inst program by typing
Inst> quit
1.6 Patch Removal Instructions
To remove a patch, use the versions remove command as you
would for any other software subsystem. The removal process
reinstates the original version of software unless you have
specifically removed the patch history from your system.
versions remove patchSGxxxxxxx
where xxxxxxx is the patch number.
To keep a patch but increase your disk space, use the
versions removehist command to remove the patch history.
versions removehist patchSGxxxxxxx
where xxxxxxx is the patch number.
INST SUBSYSTEM REQUIREMENTS No Requirements Information Available. INST SUBSYSTEM CHECKSUMS These checksums help to provide a 'signature' for the patch inst image which can be used to authenticate other inst images. You can obtain this kind of output by running sum -r on the image (from the command line): 22933 6 patchSG0002851 50196 70 patchSG0002851.eoe_man 05291 3235 patchSG0002851.eoe_sw 41313 22 patchSG0002851.idb INST SUBSYSTEM FILE LISTINGS The following lists the files which get installed from each subsystem in the patch:
DOWNLOAD PATCH
|
||||||||||||||||||||||||||||||||||||
Document Id: 20021117071318-IRIXPatch-1190
|
||||||||||||||||||||||||||||||||||||