In this Document
Installation of Oracle Clusterware 12.2.0.1 on SLES 12 SP4 fails with below error,when running root.sh
2017-05-09 14:44:51: Starting CSS in exclusive mode2017-05-09 14:44:51: Configured CRS Home: $GRID_HOME2017-05-09 14:44:51: Executing cmd: $GRID_HOME/bin/crsctl start crs -excl -cssonly2017-05-09 14:44:55: Command output:
> CRS-2672: Try to start "ora.evmd" on "<nodename1>"> CRS-2672: Try to start "ora.mdnsd" on "<nodename1>"> CRS-2676: Starting "ora.mdnsd" on "<nodename1>" successfully> CRS-2676: Starting "ora.evmd" on "<nodename1>" successfully> CRS-2672: Try to start "ora.gpnpd" on "los-oraracn1"> CRS-2676: Starting "ora.gpnpd" on "<nodename1>" successful> CRS-2672: Try to start "ora.cssdmonitor" on "<nodename1>"> CRS-2672: Try to start "ora.gipcd" on "<nodename1>"> CRS-2676: Starting "ora.cssdmonitor" on "<nodename1>" successfully> CRS-2674: Starting "ora.gipcd" on "<nodename1>" not successful> CRS-2679: Try to clean up "ora.gipcd" on "<nodename1>"> CRS-2681: Clean up of "ora.gipcd" to "<nodename1>" successfully> CRS-2673: Try to stop "ora.cssdmonitor" on "<nodename1>"> CRS-2677: Stopping "ora.cssdmonitor" on "<nodename1>" successful> CRS-2673: Try to stop "ora.mdnsd" on "<nodename1>"> CRS-2677: Stopping "ora.mdnsd" on "<nodename1>" successful> CRS-5804: Error in communication with agent process> CRS-4000: Start command not successful or completed with errors.>End Command output
2017-05-09 14:44:55: CRS-2672: Try to start "ora.evmd" on "<nodename1>"2017-05-09 14:44:55: CRS-2672: Try to start "ora.mdnsd" on "<nodename1>"2017-05-09 14:44:55: CRS-2676: Starting "ora.mdnsd" on "<nodename1>" successfully2017-05-09 14:44:55: CRS-2676: Starting "ora.evmd" on "<nodename1>" successfully2017-05-09 14:44:55: CRS-2672: Try to start "ora.gpnpd" on "<nodename1>"2017-05-09 14:44:55: CRS-2676: Starting "ora.gpnpd" on "<nodename1>" successful2017-05-09 14:44:55: CRS-2672: Try to start "ora.cssdmonitor" on "<nodename1>"2017-05-09 14:44:55: CRS-2672: Try to start "ora.gipcd" on "<nodename1>"2017-05-09 14:44:55: CRS-2676: Starting "ora.cssdmonitor" on "<nodename1>" successfully2017-05-09 14:44:55: CRS-2674: Starting "ora.gipcd" on "<nodename1>" failed successfully2017-05-09 14:44:55: CRS-2679: Try to clean up "ora.gipcd" on "<nodename1>"2017-05-09 14:44:55: CRS-2681: Clean up "ora.gipcd" on "<nodename1>" successfully2017-05-09 14:44:55: CRS-2673: Try to stop "ora.cssdmonitor" on "<nodename1>"2017-05-09 14:44:55: CRS-2677: Stopping "ora.cssdmonitor" on "<nodename1>" successfully2017-05-09 14:44:55: CRS-2673: Try to stop "ora.mdnsd" on "<nodename1>"2017-05-09 14:44:55: CRS-2677: Stopping "ora.mdnsd" on "<nodename1>" successfully2017-05-09 14:44:55: CRS-5804: Error in communication with agent process2017-05-09 14:44:55: CRS-4000: Command start not successful or completed with errors.
2017-05-09 14:44:55: The exlusive mode cluster start failed, see Clusterware alert log for more information2017-05-09 14:44:55: Executing cmd: $GRID_HOME/bin/clsecho -p has -f clsrsc -m 1192017-05-09 14:44:55: Command output:> CLSRSC-119: Start of the exclusive mode cluster failed
From GI Alert.log
2017-05-10 14: 19: 47.702 [CLSECHO (17912)] CLSRSC-0567: Beginning Oracle Grid Infrastructure configuration.2017-05-10 14: 20: 35.332 [OCRCONFIG (22865)] CRS-2101: OLR has been formatted with version 4.2017-05-10 14: 21: 17.717 [OHASD (23972)] CRS-8500: The Oracle Clusterware OHASD process starts with the operating system process ID 239722017-05-10 14: 21: 17.718 [OHASD (23972)] CRS-0714: Oracle Clusterware Release 12.2.0.1.0.2017-05-10 14: 21: 17.722 [OHASD (23972)] CRS-2112: OLR service started on node los-oraracn1.2017-05-10 14: 21: 17.730 [OHASD (23972)] CRS-1301: Oracle High Availability Service started on node los-oraracn1.2017-05-10 14: 21: 38.953 [CLSECHO (24692)] OKA-0620: OK is not supported on this operating system version: '4.4.49-92.14-default SP2'2017-05-10 14: 21: 57.605 [OHASD (25699)] CRS-8500: Oracle Clusterware's OHASD process starts with the operating system process ID 256992017-05-10 14: 21: 57.606 [OHASD (25699)] CRS-0714: Oracle Clusterware Release 12.2.0.1.0.2017-05-10 14: 21: 57.611 [OHASD (25699)] CRS-2112: OLR service started on node los-oraracn1.2017-05-10 14: 21: 57.618 [OHASD (25699)] CRS-1301: Oracle High Availability Service started on node los-oraracn1.2017-05-10 14: 21: 57.837 [ORAAGENT (25806)] CRS-8500: The ORAAGENT process of the Oracle Clusterware starts with the operating system process ID 258062017-05-10 14: 21: 57.854 [CSSDMONITOR (25832)] CRS-8500: The Oracle Clusterware CSSDMONITOR process starts with the operating system process ID 258322017-05-10 14: 21: 57.854 [CSSDAGENT (25827)] CRS-8500: Oracle Clusterware's CSSDAGENT process starts with operating system process ID 258272017-05-10 14: 21: 57.862 [ORAROOTAGENT (25819)] CRS-8500: The ORAROOTAGENT process of Oracle Clusterware starts with the operating system process ID 258192017-05-10 14: 21: 57.898 [ORAROOTAGENT (25819)] CRS-8503: The ORAROOTAGENT Oracle Clusterware process with operating system process ID 25819 received a fatal error or exception code 11.2017-05-10T14: 21: 58.064744 + 02: 00Errors in file $TRACE/ohasd_orarootagent_root.trc (incident = 985):CRS-8503 [Signal / Exception: 11] [Instruction Addr: 0x7f1ec2deb4a0] [Memory Addr: (nil)] [] [] [] [] [] [] [] []Incident details in: $INCIDENT/ohasd_orarootagent_root_i985.trc
2017-05-10 14: 21: 58.424 [ORAROOTAGENT (25878)] CRS-8500: The ORAROOTAGENT process of the Oracle Clusterware starts with the operating system process ID 258782017-05-10 14: 21: 58.455 [ORAROOTAGENT (25878)] CRS-8503: The ORAROOTAGENT Oracle Clusterware process with operating system process ID 25878 received a fatal error or exception code 11.2017-05-10T14: 21: 58.489246 + 02: 00Errors in file$TRACE/ohasd_orarootagent_root.trc (incident = 993):CRS-8503 [Signal / Exception: 11] [Instruction Addr: 0x7f91731a84a0] [Memory Addr: (nil)] [] [] [] [] [] [] [] []Incident details in: $INCIDENT/ohasd_orarootagent_root_i993.trc
2017-05-10 14: 21: 58.839 [ORAROOTAGENT (25904)] CRS-8500: The ORAROOTAGENT process of Oracle Clusterware starts with the operating system process ID 259042017-05-10 14: 22: 06.902 [ORAROOTAGENT (25904)] CRS-5822: Agent '$GRID_HOME/bin/orarootagent_root' logged off from server. Details at (: CRSAGF00117 :) {0: 11: 3} in $TRACE/ohasd_orarootagent_root.trc.[OHASD (26860)] CRS-8500: The Oracle Clusterware OHASD process starts with the operating system process ID 268602017-05-10 14: 22: 18.879 [OHASD (26860)] CRS-0714: Oracle Clusterware Release 12.2.0.1.0.2017-05-10 14: 22: 18.883 [OHASD (26860)] CRS-2112: OLR service started on node los-oraracn1.2017-05-10 14:22:18,891 [OHASD (26860)] CRS-1301: Oracle High Availability Service started on node los-oraracn1.2017-05-10 14: 22: 19.348 [ORAAGENT (26951)] CRS-8500: The ORAAGENT process of Oracle Clusterware starts with the operating system process ID 269512017-05-10 14: 22: 19.363 [CSSDAGENT (26972)] CRS-8500: Oracle Clusterware's CSSDAGENT process starts with operating system process ID 269722017-05-10 14: 22: 19.364 [CSSDMONITOR (26977)] CRS-8500: The Oracle Clusterware CSSDMONITOR process starts with the operating system process ID 269772017-05-10 14: 22: 19.372 [ORAROOTAGENT (26969)] CRS-8500: The ORAROOTAGENT process of Oracle Clusterware starts with the operating system process ID 26969[ORAROOTAGENT (26969)] CRS-8503: The Oracle clusterware process ORAROOTAGENT with operating system process ID 26969 received a fatal error or exception code 11.2017-05-10T14: 22: 19.445790 + 02: 00Errors in file $TRACE/ohasd_orarootagent_root.trc (incident = 1001):CRS-8503 [__lll_unlock_elision () + 48] [Signal / Exception: 11] [Instruction Addr: 0x7f132bc6c4a0] [Memory Addr: (n
ohasd_orarootagent_root_<incident number>.trc shows the following stack trace-
Dump file $INCIDENT/ohasd_orarootagent_root_i985.trc[TOC00000]Jump to table of contentsDump continued from file: $TRACE/ohasd_orarootagent_root.trc[TOC00001]CRS-8503 [__lll_unlock_elision()+48] [Signal/Exception: 11] [Instruction Addr: 0x7f1ec2deb4a0] [Memory Addr: (nil)] [] [] [] [] [] [] [] []
========= Dump for incident 985 (CRS 8503 [__lll_unlock_elision()+48]) ========Starting a Diag Context default dump (level=3)[TOC00003]----- START Event Driven Actions Dump -------- END Event Driven Actions Dump ----[TOC00003-END][TOC00004]----- START DDE Actions Dump -----Executing SYNC actionsExecuting ASYNC actions[TOC00005]----- START DDE Action: 'dumpFrameContext' (Sync) -----[TOC00006]----- START Frame Context DUMP -----Status: 0---------------------------------------------------------------------------[0001] (location=FILE:clsdadr.c LINE:4896 ID:dbgec*):clsdsigSignal pointer=0x7f1ebd3921e0-------------------------------------------------------------------------------- END Frame Context DUMP -----[TOC00006-END]----- END DDE Action: 'dumpFrameContext' (SUCCESS, 0 csec) -----[TOC00005-END][TOC00007]----- START DDE Action: 'dumpDiagCtx' (Sync) -----[TOC00008]------- Diag Context Dump ------------ General DiagCtx Dump -----Serial#: 0Status: 3Flags: 0x21DUsage type: 0Diag Fixed SGA information:Status: 2Flags: 0x0Incident Flag: UNSETHeap: 0x7f1ea8001170Flood Control information:Initialized: NOEnabled: NODiag init parameters:Flags: 0x1Base Flag: 0x0Options: 0xC0820ADR option: 1ADR Base: '/u01/app/oracle'Product Type: 3Product ID: 'los-oraracn1'Instance ID: 'crs'Flood Control Gate: (nil)Flood Control Ctx size: 0ADR default lib ID: 5UTS in mem trace enabled: TRUEUTS default bucket size: 4294967295UTS default bucket: 0x29fda50Incident ID cache cbf: (nil)Incident ID cache pop init: FALSEUTS default trace file attr:Flags: 0x1Prefix: 'ohasd_orarootagent_root'PIDStr: ''Suffix: ''Notify cbk: 0x7f1ec79b87feSize limits: 18446744073709551615Size cbk: (nil)DDE alive update cbk: (nil)DDE alive start cbk: (nil)DDE alive clear cbk: (nil)Custom Flood Control cbk: (nil)----- END General DiagCtx Dump ---------- DDE Diagnostic Information Dump -----Depth: 1DDE flags: 0x0Heap: 0x7f1ea8014110Incident Context pointer in diag: 0x7f1ebd38ae60Incident ID Cache: 0x7f1ea801dff0 (PGA)Invocation Context #: 0----- Invocation Context Dump -----Address: 0x7f1ea8024110Phase: 3flags: 0x10E0000Incident ID: 985Error Descriptor: CRS-8503 [__lll_unlock_elision()+48] [Signal/Exception: 11] [Instruction Addr: 0x7f1ec2deb4a0] [Memory Addr: (nil)] [] [] [] [] [] [] [] []Error class: 0Problem Key # of args: 1Number of actions: 9----- Incident Context Dump -----Address: 0x7f1ebd38ae60Incident ID: 985Problem Key: CRS 8503 [__lll_unlock_elision()+48]Error: CRS-8503 [__lll_unlock_elision()+48] [Signal/Exception: 11] [Instruction Addr: 0x7f1ec2deb4a0] [Memory Addr: (nil)] [] [] [] [] [] [] [] []
Known issue in SLES 12.glibc in SuSE 12 makes use of a Hardware Lock Elision (HLE) available in newer Intel Processors.
This exposes Bug 25851874 causing OHASD to crash on startup.
Solution 1
=================
Download fix for bug 25851874 for Linux x86-64 from the below Link.Apply this Patch to the 12.1 or 12.2 Home before running root.sh
https://updates.oracle.com/download/25851874.html
Solution 2
================
1. Assuming root.sh has already failed, deconfigure the failed install as the ROOT user:
2. Modify the /etc/ld.so.conf adding /lib64/noelision as the FIRST entry. It should look similar to the following:
3. Create a link in $GI_HOME/lib for the noelision version of the libpthread library:
4. Rerun the root.sh script and complete the installation via the OUI once root.sh has successfully completed.
Solution 3
1. Remove existing 12c Binary Installation Cleanly.
Instructions for removal of Binaries can be found here -
How to Deconfigure/Reconfigure(Rebuild OCR) or Deinstall Grid Infrastructure (Doc ID 1377349.1) ===> D. Grid Infrastructure Deinstall
2. Make the HLE related changes as below
• Create a file /etc/ld.so.conf.d/noelision.conf containing the line "/lib64/noelision".• Run "ldconfig" to rebuild caches.
3.Reinstall the GI Binaries & Retry running root.sh