Printer View

Browse Published Validated Configurations  >  Validated Configuration Details
Program Scope

The goal of this program is to validate hardware with Oracle Linux and Oracle VM, and to identify issues and configuration details that would impact customer deployments. Oracle developed the program to go far beyond the standard certification testing normally performed with operating systems. Hardware vendors participating in this program perform thorough testing of the hardware in real-world configurations with Oracle Linux and Oracle VM.

This is not a database certification. Oracle Database is a component of the environment because it is demanding on hardware resources (processor, memory, cache, storage, network) and the operating system, and well suited for this advanced level of testing. The validation program uses Oracle Database as a tool and hardware is not certified with Oracle Database as a result of the validation program.

For more information please see the program's Frequently Asked Questions (FAQ).

Validated Configuration Details
Configuration SummaryOEL 4 Update 4 x86-64 on HP ProLiant DL380 G4 and HP StorageWorks 4400 Enterprise Virtual Array with Oracle Database 10g Release 2
Publication Date14-APR-08
Server PlatformHPE ProLiant DL380 G4
Storage ModelHP StorageWorks 4400 Enterprise Virtual Array
Oracle SoftwareOracle Database 10g Release 2 ( for Linux x86_64
Linux DistributionOracle Enterprise Linux 4 Update 4 x86-64
Server and Storage Platform Details
Server Model2 X HPE ProLiant DL380 G4
Processors2 X Quad Core Xeon 2.80GHz 4M
Memory6GB RAM
Network/Interconnect1 X 10/100/1000BaseT Embedded NICs Gigabit Ethernet
HBA FCA2214 2Gb Fibre Channel HBA
Storage ModelHP StorageWorks 4400 Enterprise Virtual Array
Switch Details
Switch ModelHP Procurve 6108 1000bt
Switch TypeNetwork Switch
Additional Info.Network Switch for Interconnect
Switch Model2 X Cisco MDS9124
Switch Type2Gb 40port Fibre Channel Switch
Additional Info.FC Switch
Linux Distribution Details
OSOEL 4 Update 4 x86-64
Kernelkernel-smp-2.6.9- or higher
Additional Packages Needed From Distribution
Oracle Packages
oracleasm-support-2.0.3-1.x86_64.rpm or higherRequired for DB on ASM
oracleasmlib-2.0.2-1.x86_64.rpmRequired for DB on ASM
oracle-validated-1.0.0-4.el4.x86_64.rpm [optional, but recommended]This package automatically pulls in package dependencies and sets system parameters recommended above. It is not required but is recommended to save time in getting the system setup. It is available from ULN by subscribing to the "Oracle Software" channel and then using up2date to install it. The dependent packages will be installed and parameters will be adjusted as part of this up2date process. NOTE: For supportability and diagnosability gdb should also be installed using up2date
oracleasm-2.6.9- for DB on ASM
Configuration Files
Conf File Settings Comments
/etc/modprobe.confoptions qla2xxx ConfigRequired=1 ql2xmaxqdepth=16 qlport_down_retry=30 ql2xloginretrycount=30 ql2xfailover=1 ql2xlbType=1 ql2xautorestore=0x80 
 options hangcheck-timer hangcheck_tick=30 hangcheck_margin=180 hangcheck_reboot=1 
/etc/security/limits.conforacle hard memlock 50000000 
 oracle soft memlock 50000000# set memlock greater than or equal to the sga size to allow oracle to use hugepages if configured
 oracle soft nproc 131072 
 oracle hard core unlimited 
 oracle hard nproc 131072 
 oracle soft nofile 131072# depending on size of db, these may need to be larger
 oracle hard nofile 131072 
 oracle soft core unlimited 
 kernel.shmmax=3097020416# set to a value half the size of physical memory
  net.ipv4.ip_local_port_range = 1024 65000 
 net.core.rmem_default=262144# rmem_max can be tuned based on workload to balance performance vs lowmem usage
 kernel.sem=250 32000 100 142 
 fs.file-max = 327679 
 fs.aio-max-nr= 3145728 
Filesystems Tested
Filesystem Mount Options Details
ASMNoneUsing Automatic Storage Management Library Driver (asmlib) for datafiles
RAWNoneocr and voting disk located on raw partitions, for configuration with datafiles on ASM storage.
Driver Modules
Driver ModuleQlogic HBA driver module
Additional Info.The HBA driver module used is the version supplied with HP PSP 7.90 (Available from at the Software and Drivers download pages.)
Driver ModuleHangcheck-timer module
Additional Info.The driver used is the hangcheck-timer module supplied with the distribution
Driver ModuleNetwork driver module (tg3)
Additional Info.The driver used is the tg3 driver supplied with the distribution
Two gigabit networks, 1 for public(igb) and 1 private(igb).
The swap partition is 16GB.
Oracle Software Details
ProductOracle Database 10g Release 2, Single Instance and Oracle Real Application Clusters (RAC) for Linux x86_64
Patches5452672 , 5526245 , 6614205 , 5940792
Configuration Feedback
Bug No4593892
Summaryrepeated warnings in dmesg about lost ticks (it 80129/bz 139376)
Affectsusually seen on x86-64, both Intel and Opteron based systems
Symptomdmesg will have errors like: warning: many lost ticks. Your time source seems to be instable or some driver is hogging interrupts rip __do_softirq+0x4d/0xd0
Solution/WorkaroundFixed in U5
Bug No5041764
SummaryCFQ io scheduler can delay/hang io (bz 184535/it 88208 )
Affectsheavy io access to raw/block devices
Symptomio requests will start to hang or take a long to complete RAC node could be evicted
Solution/WorkaroundFixed in U4. Pre-U4, boot w/ elevator=deadline
Bug No5058952
Summarye1000 flow control defaults to none in the 2.6 kernel
Affectse1000 network interfaces with heavy traffic
SymptomNetwork performance of the system is affected. For e.g: In the case of a RAC cluster, this may result in lost global cache blocks.
Solution/WorkaroundFixed in U4. Pre-U4, load e1000 with FlowControl=1 (Rx)
Bug No5093059
Summarynetdump doesn't load with bonding configured (it 85145)
Affectsbringing up netdump over a bonded interface
Symptom/var/log/messages will have an error: netlog: bond0 doesn't support polling, aborting.
Solution/WorkaroundFixed in U4.
Bug No5136660
Summarybonding primary=ethX doesn't take effect
AffectsBonding doesn't take effect
SymptomIf we make a bond of eth5 and eth6, where eth6 is primary, on reboot eth5 will become active which makes server inaccessible.
Solution/WorkaroundFixed in kernel or higher
Bug No5333606
Summaryrpm throws errors if selinux is disabled at OS installation and is enabled at runtime
Affectserror at rpm packages installs
Symptomduring packages installation, errors indicate the scriplet failures
Solution/Workaroundrelabel the filesystems through 'touch /.autorelabel' or 'fixfilesrelabel'
Bug No5447805 (BZ 202186)
SummaryKernel panic on 'aio_put_req' with x86_64 RHEL 4 AS U3 when all multipathed paths fails at the same time
Affectsmultipath solutions
SymptomKernel panic on aio_out_req
Solution/WorkaroundFixed in kernel-2.6.9-42.5.EL.src.rpm (available from RedHat) or in kernel-smp- (available from Oracle)
Bug No5463714
Summarysystem will hang if reboot or restart iscsi service without unmount ocfs2 partition on iscsi volume
Affectssystem hang
Symptomrestart iscsi service or reboot without umount iscsi volume cause system hang
Solution/Workaroundunmount iscsi volume before restart iscsi service or reboot server
Bug No5593995
SummaryUsing Pro*C causes the system to run out of memory
AffectsCaused the system to hang.
SymptomThe system hangs.
Solution/WorkaroundFor OEL5U3, Remove references to pathnames that do not exist on your system in pcscfg.cfg. Fixed in OEL5U4.
Bug No5693201
SummaryApplication does not detect OEL4 as RHEL4
AffectsApplications that use /etc/redhat-release for detection of distribution
SymptomApplications that use /etc/redhat-release for detection of distribution may error out
Solution/Workaroundecho "Red Hat Enterprise Linux AS release 4 (Nahant Update 4)" > /etc/redhat-release
Bug No7240271
SummaryKernel panic occurs while starting up Oracle Enterprise Manager agent
Affects2.6.9- ( OEL ) or 2.6.9-67.0.20.EL ( RHEL ) kernels
SymptomMachine will experience a linux BUG, panic & reboot, when the EM Agent starts.
Solution/WorkaroundEL4.6 errata ( 2.6.9- and 2.6.9- kernel patches that have this fix.
Bug NoRedhat Bugzilla 202027
SummaryClear_kernel_mapping: mapping has been split. will leak memory
AffectsSeen on x86-64 Opteron based systems
Symptom/var/log/messages will have errors like clear_kernel_mapping: mapping has been split. will leak memory
Solution/WorkaroundFixed in U5
Bug No5240928
Summarywarning messages in trace files during certain workloads with async io turned on
AffectsPerformance under high i/o load
Solution/WorkaroundFixed in For 10g, versions lower than ,apply patch 5452672.
Bug No5215593
Summaryunable to start RAC with db_cache_size > 4gb
AffectsRAC instances with > 4gb db_cache_size
Symptomnodes that started up correctly with may fail to startup with ORA-4031 errors in
Solution/WorkaroundFixed in For 10g, versions lower than, apply patch 5940792; or set _ksmg_granule_size=33554432 in the init.ora. Larger sga sizes may need a higher value.
Bug No5526245
Summarywarning messages in trace files from qmnc process
Affects10.2.0.2 and RAC
SymptomFollowing warning in trace files - KSXPLOOKUP: warning cnh incarnation number may be stale ...
Solution/WorkaroundFixed in For 10g, versions lower than ,apply patch 5526245.
Bug No6057703
SummaryORA-00600:[kclchkblkdma_3] in RAC instance
Affects10.2.0.3 RAC
Symptomthe user process may terminate leaving the above mentioned traces in alert/trace files
Solution/WorkaroundFixed in For 10g, versions lower than ,apply patch 6614205.
Bug No6682370
SummaryWhen configuring the cluster interconnect the interface name should not contain a "."
Affects10.2.0.3 cluster interconnect configuration
SymptomIf the interface name being used as the private interconnect contains a "." then 1. oifcfg getif does not display the confguration; and 2. The public interface gets used as the cluster interconnect
Solution/WorkaroundFixed in For 10g, versions lower than avoid using interface names with a "."
Bug No8451866
SummaryIncorrect multipath.conf setting affects RAC node failover. Settings(features = "1 queue_if_no_path") and/or (no_path_retry = queue) causes the node hang if io path fails. no_path_retry has to be set to the numeric value explicitly.
AffectsAll applications using device mapper multipath devices which have contingency plan in case of io path failure.
SymptomRAC node hangs if io path fails.
Solution/WorkaroundIn configuration file /etc/multipath.conf,set no_path_retry and polling_interval to the numeric values, for example, "no_path_retry=5" and "polling_interval=10" so as to disable queuing. These values depend on system workload.
Bug No5768303
SummaryUnder load with device mapper the system reboots
AffectsRunning Oracle with device-mapper
Symptomunder high-stress, SCSI paths are lost and servers reboot. Messages in /var/log/messages
Solution/WorkaroundFor qla2xxx and qla2400 options qlport_down_retry=1 & ql2xretrycount=5 in modprobe.conf Upgraded qLogic driver to release 8.01.06 failback parameter IMMEDIATE set to 1 in the device mapper config file
Bug NoOracle Bugzilla 3032
AffectsThe number of paths available for i/o
Symptomwhen a host port or storage port is blocked then unblocked or when a cable is removed then restored - one or more paths are reported down in /var/log/messages a'path is down'.
Solution/WorkaroundFixed in driver, the following settings with 8.01.07 driver is required in modprobe.conf ConfigRequired=0 ql2xprocessrscn=1 ql2xprocessnotready=0 ql2xsuspendcount=3 ql2xretrycount=5 qlport_down_retry=3 ql2xfailover=0 ql2xioctltimeout=10
Note: The above configuration can be deployed with Red Hat with the changes listed below:
os Kernel
RHEL 4 AS/ES Update 4 x86-64kernel-smp-2.6.9-42.EL.x86_64.rpm or higher
Required Packages
oracleasm-support-2.0.3-1.x86_64.rpm or higher
oracleasm-2.6.9-42.ELsmp-2.0.3-1.x86_64.rpm or higher

Note: Oracle-validated for Red Hat is available from oracle-validated-1.0.0-18 onwards.

Copyright (c) 2007, Oracle Corporation. All Rights Reserved.