File contents
Copyright © 2007 Red Hat, Inc. and others. This material may be distributed
only subject to the terms and conditions set forth in the Open Publication
License, v1.0, available at http://www.opencontent.org/openpub/.
_________________________________________________________________
Introduction
The following topics are covered in this document:
* Installation-Related Notes
* Feature Updates
* Kernel-Related Updates
* Driver Updates
* Technology Previews
* Resolved Issues
* Known Issues
Some updates on Red Hat Enterprise Linux 4.7 may not appear in this version
of the Release Notes. An updated version of the Red Hat Enterprise Linux 4.7
Release Notes may also be available at the following URL:
http://www.redhat.com/docs/manuals/enterprise/
Installation-Related Notes
The following section includes information specific to installation of Red
Hat Enterprise Linux and the Anaconda installation program.
Note
When updating from one minor version of Red Hat Enterprise Linux 4 (such as
4.5 to 4.6) to Red Hat Enterprise Linux 4.7, it is recommended that you do
so using Red Hat Network, whether through the hosted web user interface or
Red Hat Network Satellite.
If you are upgrading a system with no available network connectivity, use
the "Upgrade" functionality of Anaconda. However, note that Anaconda has
limited abilities to handle issues such as dependencies on additional
repositories or third-party applications. Further, Anaconda reports
installation errors in a log file, not interactively.
As such, Red Hat recommends that when upgrading offline systems, you should
test and verify the integrity of your upgrade configuration first. Be sure
to carefully review the update log for errors before applying the upgrade to
your production environment.
In-place upgrades between major versions of Red Hat Enterprise Linux (for
example, upgrading from Red Hat Enterprise Linux 3 to Red Hat Enterprise
Linux 4.7) is not supported. While the "Upgrade" option of Anaconda allows
you to perform this, there is no guarantee that the upgrade will result in a
working installation. In-place upgrades accross major releases do not
preserve all system settings, services, and custom configurations. For this
reason, Red Hat strongly recommends that you perform a fresh installation
when planning to upgrade between major versions.
* If you are copying the contents of the Red Hat Enterprise Linux 4.7
CD-ROMs (in preparation for a network-based installation, for example)
be sure you copy the CD-ROMs for the operating system only. Do not copy
the Supplementary CD-ROM, or any of the layered product CD-ROMs, as this
will overwrite files necessary for Anaconda's proper operation.
These CD-ROMs must be installed after Red Hat Enterprise Linux is
installed.
* The version of GRUB shipped with Red Hat Enterprise Linux 4 (and all
updates) does not support software mirroring (RAID1). As such, if you
install Red Hat Enterprise Linux 4 on a RAID1 partition, the bootloader
will be installed in the first hard drive instead of the master boot
record (MBR). This will render the system unbootable.
If you wish to install Red Hat Enterprise Linux 4 on a RAID1 partition,
you should clear any pre-existing bootloader from the MBR first.
* When installing Red Hat Enterprise Linux 4 in Text Mode on systems that
use flat-panel monitors and some ATI cards, the screen area may appear
shifted. When this occurs, some areas of the screen will be obscured.
If this occurs, perform the installation with the parameter linux nofb.
* When upgrading from Red Hat Enterprise Linux 4.6 to this release,
minilogd may log several SELinux denials. These error logs are harmless,
and can be safely ignored.
Feature Updates
Password Hashing Using SHA-256/SHA-512
Password hashing using the SHA-256 and SHA-512 hash functions is now
supported.
To switch to SHA-256 or SHA-512 on an installed system, run
authconfig --passalgo=sha256 --kickstart or authconfig
--passalgo=sha512 --kickstart. Existing user accounts will not be
affected until their passwords are changed.
For newly installed systems, using SHA-256 or SHA-512 can be
configured only for kickstart installations. To do so, use the
--passalgo=sha256 or --passalgo=sha512 options of the kickstart
command auth; also, remove the --enablemd5 option if it is present.
If your installation does not use kickstart, use authconfig as
described above, then change all passwords (including root) created
after installation.
Appropriate options were also added to libuser, pam, and shadow-utils
to support these password hashing algorithms. authconfig configures
necessary options automatically, so it is usually not necessary to
modify them manually:
+ New values of the crypt_style option and new options for both
hash_rounds_min and hash_rounds_max are now supported in the
[defaults] section of /etc/libuser.conf. For more information,
refer to /usr/share/doc/libuser-[libuser version]/README.sha.
+ New options sha256, sha512, and rounds are now supported by the
pam_unix PAM module. For more information, refer to
/usr/share/doc/pam-[pam version]/txts/README.pam_unix.
+ The following new options in /etc/login.defs are now supported by
shadow-utils:
o ENCRYPT_METHOD — Specifies the encryption method to be used.
Valid values are DES, MD5, SHA256, SHA512. If this option is
defined, MD5_CRYPT_ENAB is ignored.
o SHA_CRYPT_MIN_ROUNDS and SHA_CRYPT_MAX_ROUNDS — Specifies the
number of hashing rounds to use if ENCRYPT_METHOD is set to
SHA256 or SHA512. If neither option is set, a default value is
chosen by glibc. If only one option is set, the encryption
method specifies the number of rounds.
If both options are used, they specify an inclusive interval
from which the number of rounds is chosen randomly. The
selected number of rounds is limited to the inclusive interval
[1000, 999999999].
OFED in comps.xml
The group OpenFabrics Enterprise Distribution is now included in
comps.xml. This group contains components used for high-performance
networking and clustering (for example, InfiniBand and Remote Direct
Memory Access).
Virtualization
This update implements the use of paravirtualized block device and
network drivers, which improve the performance of fully-virtualized
guests. In addition, you can now use more than three virtual network
interface (VNIF) numbers per guest domain.
divider
The divider=[value] option is a kernel command-line parameter that
allows you to adjust the system clock rate while maintaining the same
visible HZ timing value to user space applications.
Using the divider=[value] option allows you to reduce CPU overhead
and increase efficiency at the cost of lowering the accuracy of
timing operations and profiling. This is useful in virtualized
environments as well as for certain applications.
Useful [values] for the standard 1000Hz clock are:
+ 2 = 500Hz
+ 4 = 250Hz
+ 10 = 100Hz (value used by previous releases of Red Hat Enterprise
Linux)
Note that the virtualized kernel uses a 250HZ clock by default. As
such, it does not need the divider=[value] option either in dom0 or
in paravirtualized guests.
Firefox Rebase
Firefox is now updated to version 3.0. This update features several
fixes and enhancements, most notably:
+ Set homepages are now loaded correctly when the Firefox browser
window is opened.
+ Firefox no longer crashes when you search for the string "do".
+ Firefox in 64-bit mode now loads the ext JavaScript library
correctly. In previous versions of Firefox, web-based applications
that used this library either took too long to load, or were never
loaded at all.
+ A cross-site scripting flaw was discovered in the way Firefox
handled the jar:URI scheme. This flaw made it possible for a
malicious web site to conduct a scripting attack against the user.
This security issue is now fixed in this update.
+ Several flaws were discovered in the way Firefox processed certain
malformed content. Web sites that contained such content could
cause Firefox to crash or even execute arbitrary code as the user
running Firefox. This security issue is now fixed in this update.
+ A race condition was discovered in the way Firefox set the
window.location property on a web page. With this flaw, it was
possible for a web page to set an arbitrary Referer header; this
could lead to a cross-site request forgery (CSRF) attack against
websites that rely only on the Referer header. This security issue
is now fixed in this update.
+ Firefox now renders correctly on laptops equipped with external
display.
Note, however, that this update of Firefox is not fully backwards
compatible with all JavaScripts or Firefox plugins used today.
Also, Red Hat has observed that several large commercial web
applications have relied on the presence of some cross-site scripting
flaws addressed by this Firefox update. These scripting flaws are
described in the following links:
+ http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2008-1234
+ http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2008-0415
Consequently, the use of these commercial web applications may result
in some loss of functionality. You can observe this in the presence
of additional JavaScript errors in the Firefox Error Console (Tools
=> Error Console). Red Hat is currently working with the
corresponding vendors to address this.
Kernel-Related Updates
General Kernel Updates
+ iostat now outputs statistics regarding the status and I/O
performance of partitions.
+ I/O accounting in this release now outputs more comprehensive core
statistics. This was accomplished by implementing the use of
ru_inblock and ru_outblock, which are already used upstream.
+ show_mem() output now includes the total number of pagecache pages.
This makes debugging information sent to the console and to
/var/log/messages more useful, particularly during out-of-memory
kills.
+ The O_ATOMICLOOKUP flag is now removed. This flag is not used by
any current userspace daemons. Further, the bit normally used by
O_ATOMICLOOKUP is used by another flag (O_CLOEXEC); as such,
O_ATOMICLOOKUP was removed to avoid any conflicts arising from this
bit share.
+ The kernel now exports process limit information to
/proc/[PID]/limits (where [PID] is the process ID).
+ The parameter TCP_RTO_MIN can now be configured to a maximum of
3000 milliseconds. TCP_RTO_MIN was not a tunable kernel parameter
in previous releases.
This update allows more TCP/IP flexibility, and enables
applications to restart a transmission in accordance with wireless
transmissions (for example, mobile phone transmission rates).
You can configure the TCP_RTO_MIN parameter through ip route. For
example, to set TCP_RTO_MIN to the maximum of 3000 milliseconds,
use:
ip route change [route] dev eth0 rto_min 3s
For more information about ip route, refer to man ip.
+ The udp_poll() function is now implemented. This update reduces the
likelihood of false positive returns from the system call select().
+ You can now enable/disable 32-bit inode numbers. To do so, use the
kernel parameter nfs.enable_ino64=. Setting nfs.enable_ino64=0 will
instruct the NFS client to return 32-bit inode numbers for
readdir() and stat() system calls (instead of the full 64-bit inode
numbers).
By default, this kernel parameter is set to return the actual
64-bit inode numbers.
+ You can now restrict NFS writes to low memory. To do so, set
/proc/sys/vm/nfs-writeback-lowmem-only to 1 (this is set to 0 by
default).
Previous releases did not include this capability. This caused NFS
read performance degradation in some cases, particularly when the
system encountered high volumes of NFS read/write requests.
+ You can now set whether mapped file pages are used in dirty_ratio
and dirty_background_ratio calculations. To do so, set
/proc/sys/vm/write-mapped to 1 (this is set to 0 by default).
Setting /proc/sys/vm/write-mapped to 1 allows you to implement
faster NFS read performance. Note, however, that doing so exposes
you to out-of-memory risks.
+ CIFS is now updated to version 1.50c. This update applies several
enhancements and bug fixes, including the capability to mount OS/2
shares.
+ Core dump masking is now supported. This allows a core dump process
to skip the shared memory segments of a process when creating a
core dump file. This feature also allows you to select whether or
not to dump anonymous shared memory for each process.
When a process is dumped, all anonymous memory is written to a core
file as long as the size of the core file isn't limited. In some
cases, you may want to prevent some memory segments (such as huge
shared memory) from being dumped. Conversely, you may also want to
save file-backed memory segments into a core file, in addition to
individual files.
For these purposes, you can use /proc/[pid]/coredump_filter to
specify which memory segments of the [pid] process is dumped.
coredump_filter is a bitmask of memory types. If a bitmask is set,
memory segments of the corresponding memory type are dumped.
The following memory types are supported:
o bit 0 — anonymous private memory
o bit 1 — anonymous shared memory
o bit 2 — file-backed private memory
o bit 3 — file-backed shared memory
To set a bitmask for [pid], set the corresponding bitmask to
/proc/[pid]/coredump_filter. For example, to prevent a dump of all
shared memory segments attached to process 1111, use:
echo 0x1 > /proc/1111/coredump_filter
The default value of coredump_filter is 0x3, which specifies that
all anonymous memory segments are dumped. Also, note that
regardless of the bitmask status, MMIO pages (such as frame
buffers) are never dumped and vDSO pages are always dumped
When a new process is created, the process inherits the bitmask
status from its parent. As such, Red Hat recommends that you set up
coredump_filter before the program runs. To do so, echo the desired
bitmask to /proc/self/coredump_filter before running the program.
Kernel Updates Specific To This Platform
+ Oprofile now supports event-based profiling on Greyhound hardware.
+ AMD ATI SB800 SATA controller is now supported.
+ AMD ATI SB600 and SB700 SATA controllers that use the 40-pin IDE
cable are now supported.
+ 64-bit direct memory access (DMA) is now supported on the AMD ATI
SB700.
+ The PCI device IDs necessary for supporting Intel ICH10 have been
added.
Driver Updates
General Driver/Platform Updates
+ i2c-piix4 kernel module is now enabled to support the AMD SBX00
SMBus.
+ i5000_edac: driver added to support Intel 5000 chipsets.
+ i3000_edac: driver added to support Intel 3000 3010 chipsets.
+ Correct cache info of Intel Tolapai chipset is now added. This
ensures that the hardware is listed correctly.
+ wacom: driver updated to add support for the following input
devices:
o Intuos3 12x19
o Intuos3 12x12
o Intuos3 4x6
o Cintiq 20wsx
+ i2c-i801: driver (along with corresponding PCI IDs) updated to
support Intel Tolapai.
+ sata_svw: driver updated to support Broadcom HT1100 chipset.
+ libata: driver updated to enable Native Command Queuing (NCQ) by
removing Hitachi drives from blacklist.
+ ide: driver updated to include ide=disable, a kernel PCI module
parameter that can be used to disable ide drivers.
+ psmouse: driver updated to properly support input devices that use
cortps protocol. Examples of these input devices are 4-button mice
and trackball devices developed by Cortron.
+ eHEA: driver updated to match upstream version. This update applies
several upstream bug fixes and enhancements that improve support
for IBM i6 and p6, including:
o Addition of Large Receive Offload (LRO) support as a
networking module.
o Addition of poll_controller, necessary for supporting netdump
and netconsole modules.
+ zfcp: driver updated to apply upstream bug fixes. This update
applies several bug fixes, most notably:
o When adapters are reopened in a multipath environment after a
fibre-channel hot-removal, affected paths are no longer marked
as failed. With this update, relevant adapter flags are now
correctly cleared during such an event.
o When an fsf request times out, the adapter is no longer marked
as failed after a successful recovery. The
ZFCP_STATUS_COMMON_ERP_FAILED flag is now cleared when the
adapter is successfully enabled.
o The BOXED flag is now cleared when the adapter is successfully
re-enabled.
o A bug that could cause a deadlock between the SCSI stack and
ERP thread (in some cases, when some devices are being
registered) is now fixed.
o When using chccwdev to mark a device as "offline" in a
multipath environment, I/O no longer stalls on all paths. In
addition, when using chccwdev to bring the same device back
online, it will still use the correct path checker.
Network
+ bnx2x: driver added to support network adapters on Broadcom 5710
chipset.
+ cxgb3: driver updated to support Chelsio 10G ethernet controller
and OFED.
+ realtek: driver updated to support Realtek RTL8111 and RTL8168
PCI-E network interface card.
+ e1000: driver updated to support alternate MAC addresses, necessary
for supporting the Virtual Connect architecture.
+ e1000e: driver updated to latest upstream version. This update
provides support for ICH9m and 82574L Shelter Island network
interface cards, and applies several upstream fixes as well.
+ bnx2: driver updated to version 1.6.9. This applies several
upstream changes, and provides support for the Broadcom 5709s
chipset.
+ igb: driver updated to upstream version 1.0.8-k2. This driver
version now supports the Intel 82575EB (Zoar) chipset.
+ s2io: driver updated to version 2.0.25.1 to support Neterion
Xframe-II 10GbE network adapter.
+ tg3: driver updated to upstream version 3.86. This update applies
several fixes and enhancements, including:
o An irq_sync race condition issue is now fixed.
o Auto-MDI is now enabled.
+ forcedeth: driver updated to upstream version 0.61. This update
provides support for the following chipsets:
o MCP73
o MCP77
o MCP79
This update also provides several bug fixes related to WOL, MAC
address ordering, and tx timeout issues.
Storage
+ stex: driver updated to version 3.6.0101.2. This update applies
several upstream enhancements and bug fixes.
+ mpt fusion: driver family updated to version 3.12.19.00. This
update applies several enhancements and fixes, most notably:
o Tuning parameters for modifying queue depth are now included
in mptsas.c, mptspi.c, and mptfc.c. These parameters are
mptsas_device_queue_depth, mptspi_device_queue_depth, and
mptfc_device_queue_depth. The default value for these tuning
parameters is 48.
o On systems with more than 36GB of memory, up to 1,078
scatter/gather entries are now supported.
o Added the flag ioc->broadcast_aen_busy. This flag is set when
the mptsas_broadcast_primative_work thread is running. When
additional aen events are posted, they will be ignored while
the ioc->broadcast_aen_busy flag is set.
In addition, SCSI_IO commands will be frozen and re-queued
later when the ioc->broadcast_aen_busy flag is set. This flag
is cleared once the mptsas_broadcast_primative_work thread is
completed.
o The internal command timeout routine now issues a Diagnostic
Reset to clear a hang condition when a sync cache command is
issued when a driver is unloaded. This capability was added by
implementing two timeout routines: one timeout routine handles
all internal commands not associated with domain validation,
while the other only handles commands associated with domain
validation.
o The domain validation timeout routine now issues a bus reset
rather than a target reset.
o When a Task Management request (via an IOCTL interface)
completes, the associated IOCTL timer is now deleted. This
prevents the timer from expiring; timer expiration results in
a host reset even when the Task Management request coompletes
successfully.
+ qla2xxx: driver updated to version 8.02.00-k5. This update to
qla2xxx adds support for the following:
o EHAFT, a QLogic host bus adapter mechanism that provides
activity information about fibre channel devices.
o 8GB fibre-channel devices.
This update also applies several improvements provided from
upstream, including a bug fix that prevents the driver from waiting
on a loop that is already in a LOOP_DEAD state.
+ qla3xxx: driver updated to version v2.03.00-k4-rhel4.7-01. This
update fixes a bug wherein if an interface owned by the qla3xxx
driver was used in a VLAN, inbound completions were not handled and
passed to the TCP/IP stack.
+ qla4xxx: driver updated to version 5.01.03-d0. This applies the
following fixes:
o A session is now created for each available port on the same
target. In addition, a bug that prevented qla4xxx from issuing
a re-login request for some targets (after a failover or cable
pull) is now fixed.
o In previous versions of qla4xxx, I/O queue depths were
unaffected by "queue full" errors. Now, queue depths are
adjusted appropriately when "queue full" errors occur, which
improves I/O error handling. Consequently, the queue depth for
all LUNs on each target will be limited as well.
o The SCSI function is now enabled before firmware
initialization. This bug fix ensures that the SCSI function
receives notification of any soft resets or fatal errors that
occur before the firmware initialization is completed.
o A bug that prevented the operating system from scanning some
targets that transition from "non-active" to "active" state
(during driver initialization) is now fixed.
+ CCISS: driver updated to version to 3.6.20-RH1. This update
provides support for upcoming SAS/SATA controllers, and applies the
following changes as well (among others):
o I/O control sg_io added. This ioctl is provided to enhance
support for multipathing.
o /proc/driver/cciss entries have been modified to prevent
system crashes when a large number of drives are installed on
the system.
o The READ_AHEAD setting in the cciss driver is now removed. The
cciss driver will now use the block layer default of 256.
Testing has shown that the setting READ_AHEAD=1024 did not
result in a consistent improvement in performance; in some
situations, this setting could also cause the system to hang.
+ megaraid_sas: driver updated to version 3.18 to support LSI 1078
chipsets running in MegaRAID mode. In addition, several bug fixes
are also applied by this update, including:
o MFI_POLL_TIMEOUT_SECS is now 60 seconds (increased from 10
seconds). This was done to accommodate the firmware, which
could take a maximum of 60 seconds to respond to the INIT
command.
o A bug that caused continuous chip resets and command timeouts
due to frame count calculation is now fixed. With this update,
the driver now sends the correct frame count to the firmware
upon request.
o Added module parameter poll_mode_io to support polling.
+ arcmsr: driver updated to version 1.20.00.15.rh. This update
applies several bug fixes and minor enhancements; in addition, it
also provides support for the following SATA RAID adapters:
o ARC1200
o ARC1201
o ARC1202
Technology Previews
Technology Preview features are currently not supported under Red Hat
Enterprise Linux 4.7 subscription services, may not be functionally
complete, and are generally not suitable for production use. However, these
features are included as a customer convenience and to provide the feature
with wider exposure.
Customers may find these features useful in a non-production environment.
Customers are also free to provide feedback and functionality suggestions
for a technology preview feature before it becomes fully supported. Erratas
will be provided for high-severity security issues.
During the development of a technology preview feature, additional
components may become available to the public for testing. It is the
intention of Red Hat to fully support technology preview features in a
future release.
Systemtap
Systemtap provides free software (GPL) infrastructure to simplify the
gathering of information about the running Linux system. This assists
diagnosis of a performance or functional problem. With the help of
systemtap, developers no longer need to go through the tedious and
disruptive sequence of instrument, recompile, install, and reboot
that may be otherwise required to collect data.
gcc
The GNU Compiler Collection (gcc-4.1) is still included in this
release as a Technology Preview. This compiler was originally
introduced as a Technology preview in Red Hat Enterprise Linux 4.4.
For more information about gcc-4.1, refer to the project website at
http://gcc.gnu.org/. An in-depth manual for gcc-4.1.2 can also be
read at http://gcc.gnu.org/onlinedocs/gcc-4.1.2/gcc/.
OpenOffice 2.0
OpenOffice 2.0 is now included in this release as a Technology
Preview. This suite features several improvements, including ODF and
PDF functionalities, support for digital signatures and greater
compatibility with open suites in terms of format and interface. In
addition to this, the OpenOffice 2.0 spreadsheet has enhanced pivot
table support, and can now handle up to 65,000 rows.
For more information about OpenOffice 2.0, please refer to
http://www.openoffice.org/dev_docs/features/2.0/index.html.
autofs5
autofs5 is included in this release as a Technology Preview. This new
version of autofs resolves several long-standing interoperability
issues in multi-vendor environments. autofs5 also features the
following enhancements:
+ direct map support, which provides a mechanism to automatically
mount file systems at any point in the file system heirarchy
+ lazy mount and umount support
+ enhanced LDAP support through a new configuration file,
/etc/autofs_ldap_auth.conf
+ complete implementation of nsswitch.conf use
+ multiple master map entries for direct maps
+ complete implementation of map inclusion, which allows the contents
of specified maps to be included in autofs master maps
At present, the autofs5 master map lexical analyzer cannot properly
parse quoted strings in the mount point or map specification. As
such, quoted strings should be written in the maps themselves.
autofs is still installed and run by default in this update. As such,
you need to install the autofs5 package manually if you wish to use
autofs5 enhancements.
You can have both autofs and autofs5 installed. However, only one of
them should be used to provide automount services. To install autofs5
and use it as your automounter, perform these steps:
1. Log in as root and stop the autofs service using the command
service autofs stop.
2. Disable the autofs service using the command chkconfig autofs off.
3. Install the autofs5 package.
4. Enable the autofs5 service using the command chkconfig autofs5 on.
5. Start autofs5 using the command service autofs5 start.
For more information about autofs5, refer to the following man pages
(after installing the autofs5 package):
+ autofs5(5)
+ autofs5(8)
+ auto.master.v5(5)
+ automount5(8)
You can also consult
/usr/share/doc/autofs5-<version>/README.v5.release for more
information.
Resolved Issues
* When an application such as systool reads /sys/class/scsi_host/host<scsi
host number>/mbox (generated by the Emulex lpfc driver), the benign "Bad
State" message is no longer printed in the console or logged into the
system log file.
* The kernel now asserts Data Terminal Ready (DTR) signals before printing
to serial ports during boot time. DTR assertion is required by some
devices. Kernel boot messages are now printed to serial consoles on such
devices.
* In Red Hat Enterprise Linux 4.6, the login prompt may not appear when
the operating system was installed through a serial console. This issue
is now fixed in this release.
Known Issues
* A bug in previous versions of openmpi and lam may prevent you from
upgrading these packages. This same bug may cause up2date to fail when
upgrading all packages.
This bug manifests in the following error when attempting to upgrade
openmpi or lam:
error: %preun(openmpi-[version]) scriptlet failed, exit status 2
This bug also manifests in the following error (logged in
/var/log/up2date) when attempting to upgrade all packages through
up2date:
up2date Failed running rpm transaction - %pre %pro failure ?.
As such, you need to manually remove older versions of openmpi and lam
first in order to avoid these errors. To do so, use the following rpm
command:
rpm -qa | grep '^openmpi-\|^lam-' | xargs rpm -e --noscripts
--allmatches
* When a LUN is deleted on a configured storage system, the change is not
reflected on the host. In such cases, lvm commands will hang
indefinitely when dm-multipath is used, as the LUN has now become stale.
To work around this, delete all device and mpath link entries in
/etc/lvm/.cache specific to the stale LUN. To find out what these
entries are, run the following command:
ls -l /dev/mpath | grep <stale LUN>
For example, if <stale LUN> is 3600d0230003414f30000203a7bc41a00, the
following results may appear:
lrwxrwxrwx 1 root root 7 Aug 2 10:33 /3600d0230003414f30000203a7bc41a00 -> ../
dm-4
lrwxrwx--rwx 1 root root 7 Aug 2 10:33 /3600d0230003414f30000203a7bc41a00p1 ->
../dm-5
This means that 3600d0230003414f30000203a7bc41a00 is mapped to
two mpath links: dm-4 and dm-5.
As such, the following lines should be deleted from /etc/lvm/.cache:
/dev/dm-4
/dev/dm-5
/dev/mapper/3600d0230003414f30000203a7bc41a00
/dev/mapper/3600d0230003414f30000203a7bc41a00p1
/dev/mpath/3600d0230003414f30000203a7bc41a00
/dev/mpath/3600d0230003414f30000203a7bc41a00p1
* In a HA-RAID two-system configuration, two SAS adapters are plugged in
to two systems and connected to a shared SAS disk drawer. Setting the
Preferred Dual Adapter State attribute to Primary on both SAS adapters
can trigger a race condition and cause infinite failover between the two
SAS adapters. This is because only one SAS adapter can be set to
Primary.
To prevent this error, ensure that the Preferred Dual Adapter State of
one SAS adapter is set to None if the other SAS adapter should be set to
Primary.
* If you need to use the hp_sw kernel module, install the updated
device-mapper-multipath package.
You also need to properly configure the HP array to correctly use
active/passive mode and recognize connections from a Linux machine. To
do this, perform the following steps:
1. Determine what the world wide port name (WWPN) of each connection
is using show connections. Below is a sample output of show
connections on an HP MSA1000 array with two connections:
Connection Name: <Unknown>
Host WWNN = 200100E0-8B3C0A65
Host WWPN = 210100E0-8B3C0A65
Profile Name = Default
Unit Offset = 0
Controller 2 Port 1 Status = Online
Connection Name: <Unknown>
Host WWNN = 200000E0-8B1C0A65
Host WWPN = 210000E0-8B1C0A65
Profile Name = Default
Unit Offset = 0
Controller 1 Port 1 Status = Online
2. Configure each connection properly using the following command:
add connection [connection name] WWPN=[WWPN ID] profile=Linux
OFFSET=[unit offset]
Note that [connection name] can be set arbitrarily.
Using the given example, the proper commands should be:
add connection foo-p2 WWPN=210000E0-8B1C0A65 profile=Linux OFFSET=0
add connection foo-p1 WWPN=210100E0-8B3C0A65 profile=Linux OFFSET=0
3. Run show connections again to verify that each connection is
properly configured. As per the given example, the correct
configuration should be:
Connection Name: foo-p2
Host WWNN = 200000E0-8B1C0A65
Host WWPN = 210000E0-8B1C0A65
Profile Name = Linux
Unit Offset = 0
Controller 1 Port 1 Status = Online
Connection Name: foo-p1
Host WWNN = 200100E0-8B3C0A65
Host WWPN = 210100E0-8B3C0A65
Profile Name = Linux
Unit Offset = 0
Controller 2 Port 1 Status = Online
* Red Hat discourages the use of quota on EXT3 file systems. This is
because in some cases, doing so can cause a deadlock.
Testing has revealed that kjournald can sometimes block some
EXT3-specific callouts that are used when quota is running. As such, Red
Hat does not plan to fix this issue in Red Hat Enterprise Linux 4, as
the modifications required would be too invasive.
Note that this issue is not present in Red Hat Enterprise Linux 5.
* Hardware testing for the Mellanox MT25204 has revealed that an internal
error occurs under certain high-load conditions. When the ib_mthca
driver reports a catastrophic error on this hardware, it is usually
related to an insufficient completion queue depth relative to the number
of outstanding work requests generated by the user application.
Although the driver will reset the hardware and recover from such an
event, all existing connections at the time of the error will be lost.
This generally results in a segmentation fault in the user application.
Further, if opensm is running at the time the error occurs, then you
need to manually restart it in order to resume proper operation.
* The Desktop Sharing connection icon displays its context menu when you
double-click it, not when you right-click it. All other icons display
their context menus when you right-click on them.
( x86 )