Solaris Tips : CPU Cache Sizes, Changing System Date

Tip #1: Finding the CPU cache sizes from Solaris operating environment

Use the prtpicl utility to list out system configuration, and look for the cache sizes within that output.

eg.,

$ prtpicl -v |grep cache
              :l1-icache-size    0x10000
              :l1-icache-line-size       0x40
              :l1-icache-associativity   0x2
              :l1-dcache-size    0x10000
              :l1-dcache-line-size       0x40
              :l1-dcache-associativity   0x2
              :l2-cache-size     0x500000
              :l2-cache-line-size        0x100
              :l2-cache-associativity    0xa

[Updated 01/14/13] The above output was gathered from an M4000 system that has SPARC64 VII processors.

Recent update releases of Solaris 10 and 11 show the prtpicl reported cache sizes in decimal numbers.

Here is a slightly improved prtpicl command that filters out unwanted output. (Courtesy: Georg)

/usr/sbin/prtpicl -v -c cpu | egrep "^ +cpu|ID|cache"

Tip #2: Changing the System Date

Use date to change the system date. For example, to set the system date to March 9, 2008 08:15 AM, run the following command. Syntax: date mmddHHMMyy

#date 0309081508

Sun Mar 9 08:15:03 PST 2008

Read More

emca fails with "Database instance is unavailable" though available

The following example shows the symptoms of failure, and the exact error message.

$ emca -repos create

...
Password for SYSMAN user:  

Do you wish to continue? [yes(Y)/no(N)]: Y
Nov 19, 2012 10:33:42 AM oracle.sysman.emcp.DatabaseChecks checkDbAvailabilityImpl
WARNING: ORA-01034: ORACLE not available

Nov 19, 2012 10:33:42 AM oracle.sysman.emcp.DatabaseChecks throwDBUnavailableException
SEVERE: 
Database instance is unavailable. Fix the ORA error thrown and run EM Configuration Assistant again.

Some of the possible reasons may be : 

1) Database may not be up. 
2) Database is started setting environment variable ORACLE_HOME with trailing '/'. Reset ORACLE_HOME and bounce the database. 

For eg. Database is started setting environment variable ORACLE_HOME=/scratch/db/ . Reset ORACLE_HOME=/scratch/db  and bounce the database.

Fix:

Ensure that the ORACLE_HOME is pointing to the right location in $ORACLE_HOME/bin/emca file.

Rather than installing from scratch, if ORACLE_HOME was copied over from another location, likely it results in wrong location for ORACLE_HOME in several Enterprise Manager (EM) specific scripts and files. It usually happens when the directory structure on the target machine is not identical to the structure on the original/source machine, including the top level directory location where Oracle RDBMS was installed properly using the installer.

Read More

Consolidating Oracle E-Business Suite R12 on Oracle's SPARC SuperCluster

An Optimized Solution for Oracle E-Business Suite (EBS) R12 12.1.3 is now available on oracle.com.

    The Oracle Optimized Solution for Oracle E-Business Suite

This solution was centered around the engineered system, SPARC SuperCluster T4-4. Check the business and technical white papers along with a bunch of relevant useful resources online at the above optimized solution page for EBS.

What is an Optimized Solution?

Oracle's Optimized Solutions are designed, tested and fully documented architectures that are tuned for optimal performance and availability. Optimized solutions are NOT pre-packaged, fully tuned, ready-to-install software bundles that can be downloaded and installed. An optimized solution is usually a well documented architecture that was thoroughly tested on a target platform. The technical white paper details the deployed application architecture along with various observations from installing the application on target platform to its behavior and performance in highly available and scalable configurations.

Oracle E-Business Suite R12 Use Case

Multiple E-Business Suite R12 12.1.3 application modules were tested in this optimized solution -- Financials (online - oracle forms & web requests), Order Management (online - oracle forms & web req uests) and HRMS (online - web requests & payroll batch). The solution will be updated with additional application modules, when they are available.

Oracle Solaris Cluster is responsible for the high availability portion of the solution.

Performance Data

For the sake of completeness, test results were also documented in the optimized solution white paper. Those test results are mainly for educational purposes only. They give good sense of application behavior under the circumstances the application was tested. Since the major focus of the optimized solution is around highly available and scalable configurations, the application was configured to me et those criteria. Hence the documented test results are not directly comparable to any other E-Business Suite performance test results published by any vendor including Oracle. Such an attempt may lead to skewed, incorrect conclusions.

Questions & Requests

Feel free to direct your questions to the author of the white papers. If you are a potential customer who would like to test a specific E-Business Suite application module on any non-engineered syste m such as SPARC T4-X or engineered system such as SPARC SuperCluster, contact Oracle Solution Center.

Read More

E-Business Suite : Role of CHUNK_SIZE in Oracle Payroll

Different batch processes in Oracle Payroll flow have the ability to spawn multiple child processes (or threads) to complete the work in hand. The number of child processes to fork is controlled by the THREADS parameter in APPS.PAY_ACTION_PARAMETERS view.

THREADS parameter

The default value for THREADS parameter is 1, which is fine for a single-processor system but not optimal for the modern multi-core multi-processor systems. Setting the THREADS parameter to a value equal to or less than the total number of [virtual] processors available on the system may improve the performance of payroll processing. However on the down side, since multiple child processes operate against the same set of payroll tables in HR schema, database may experience undesired consequences such as buffer busy waits and index contention, which results in giving up some of the gains achieved by using multiple child processes/threads to process the work. Couple of other action parameters, CHUNK_SIZE and CHUNK_SHUFFLE, help alleviate the database contention.

eg.,

Set a value for THREADS parameter as shown below.

CONNECT APPS/APPS_PASSWORD

UPDATE PAY_ACTION_PARAMETERS
SET PARAMETER_VALUE = DESIRED_VALUE
WHERE PARAMETER_NAME = 'THREADS';

COMMIT;

(I am not aware of any maximum value for THREADS parameter)

CHUNK_SIZE parameter

The size of each commit unit for the batch process is controlled by the CHUNK_SIZE action parameter. In other words, chunking is the act of splitting the assignment actions into commit groups of desired size represented by the CHUNK_SIZE parameter. The default value is 20, and each thread processes one chunk at a time -- which means each child process inserts or processes 20 assignment actions at any time.

When multiple threads are configured, each thread picks up a chunk to process, completes the assignment actions and then picks up another chunk. This is repeated until all the chunks are exhausted.

It is possible to use different chunk sizes in different batch processes. During the initial phase of processing, CHUNK_SIZE number of assignment actions are inserted into relevant table(s). When multiple child processes are inserting data at the same time into the same set of tables, as explained earlier, database may experience contention. The default value of 20 is mostly optimal in such a case. Experiment with different values for the initial phase by +/-10 for CHUNK_SIZE parameter and observe the performance impact. A larger value may make sense during the main processing phase. Again experimentation is the key in finding the suitable value for your environment. Start with a large value such as 2000 for the chunk size, then increment or decrement the size by 500 at a time until an optimal value is found.

eg.,

Set a value for CHUNK_SIZE parameter as shown below.

CONNECT APPS/APPS_PASSWORD

UPDATE PAY_ACTION_PARAMETERS
SET PARAMETER_VALUE = DESIRED_VALUE
WHERE PARAMETER_NAME = 'CHUNK_SIZE';

COMMIT;

CHUNK_SIZE action parameter accepts a value that is as low as 1 or as high as 16000.

CHUNK SHUFFLE parameter

By default, chunks of assignment actions are processed sequentially by all threads - which may not be a good thing especially given that all child processes/threads performing similar actions against the same set of tables almost at the same time. By saying not a good thing, I mean to say that the default behavior leads to contention in the database (in data blocks, for example).

It is possible to relieve some of that database contention by randomizing the processing order of chunks of assignment actions. This behavior is controlled by the CHUNK SHUFFLE action parameter. Chunk processing is not randomized unless explicitly configured.

eg.,

Set chunk shuffling as shown below.

CONNECT APPS/APPS_PASSWORD

UPDATE PAY_ACTION_PARAMETERS
SET PARAMETER_VALUE = 'Y'
WHERE PARAMETER_NAME = 'CHUNK SHUFFLE';

COMMIT;

Finally I recommend checking the following document out for additional details and additional pay action tunable parameters that may speed up the processing of Oracle Payroll.
    My Oracle Support Doc ID: 226987.1 Oracle 11i & R12 Human Resources (HRMS) & Benefits (BEN) Tuning & System Health Checks

Also experiment with different combinations of parameters and values until the right set of action parameters and values are found for your deployment.

Read More

Enabling 2 GB Large Pages on Solaris 10

Few facts:

• - 8 KB is the default page size on Solaris 10 and 11 as of this writing
  
• - both hardware and software must have support for 2 GB large pages
  
• - SPARC T4 hardware is capabile of supporting 2 GB pages
  
• - Solaris 11 kernel has in-built support for 2 GB pages
  
• - Solaris 10 has no default support for 2 GB pages
  
• - Memory intensive 64-bit applications may benefit the most from using 2 GB pages
  

Prerequisites:

OS: Solaris 10 8/11 (Update 10) or later
Hardware: SPARC T4. eg., SPARC T4-1, T4-2 or T4-4

Steps to enable 2 GB large pages on Solaris 10:

1. Install the latest kernel patch or ensure that 147440-04 or later was installed
   

   • Check the patch download instructions

2. Add the following line to /etc/system and reboot
   

   • set max_uheap_lpsize=0x80000000

3. Finally check the output of the following command when the system is back online
   

   • pagesize -a

   eg.,

   % pagesize -a
   8192  <-- 8K
   65536  <-- 64K
   4194304  <-- 4M
   268435456 <-- 256M
   2147483648 <-- 2G
   
   % uname -a
   SunOS jar-jar 5.10 Generic_147440-21 sun4v sparc sun4v
   

Also See:

• Solaris 9 or later: More performance with Large Pages (MPSS)
  
• Large page support for instructions (text) in Solaris 10 1/06
  
• Solaris: How To Disable Out Of The Box (OOB) Large Page Support?
  
• Memory fragmentation / Large Pages on Solaris x86
  

Read More

[OID] ldap_modify: Failed to find member in mandatory or optional attribute list

A sample LDAP entry and the resulting error message are shown below. The objective is simple - adding a new member (employee) to an existing group (Administrators).

% cat assigngrp.ldif

dn: cn=Administrators,ou=groups,ou=entapp
changetype: modify
add: member
member: cn=emp1234,ou=people,ou=entapp

% ldapmodify -p 3060 -h localhost -D "cn=orcladmin" -w passwd -f assigngrp.ldif
add member:
 cn=emp1234,ou=people,ou=entapp
modifying entry cn=Administrators,ou=groups,ou=entapp
ldap_modify: Object class violation
ldap_modify: additional info: Failed to find member in mandatory or \
     optional attribute list.

The above error message is a generic one. It would have been nice had it shown the expected and actual inputs as part of the error. However it gave us a hint that the object class was violated. In this example, the group "Administrators" was created under object class groupOfUniqueNames.

% ldapsearch -p 3060 -h localhost -b "ou=groups,ou=entapp" -A "(objectclass=*)"
..
cn=Administrators,ou=groups,ou=entapp
Administrators,groups,entapp
cn
uniquemember
objectclass
..

RFC 4519 for Lightweight Directory Access Protocol (LDAP) requires the uniqueMember attribute within the groupOfUniqueNames object class. An excerpt from the original RFC:

3.6.  'groupOfUniqueNames'
        ...

      ( 2.5.6.17 NAME 'groupOfUniqueNames'
         SUP top
         STRUCTURAL
         MUST ( uniqueMember $
               cn )
         MAY ( businessCategory $
               seeAlso $
               owner $
               ou $
               o $
               description ) )

Going back to the issue in hand, the "add" attribute must be uniqueMember, not member, in "modify" LDAP entry. That's the object class violation in this case. Now the fix to the issue is obvious.

The modified entry and the output from Oracle Internet Directory's ldapmodify command are shown below.

% cat assigngrp.ldif

dn: cn=Administrators,ou=groups,ou=entapp
changetype: modify
add: uniqueMember
uniqueMember: cn=emp1234,ou=people,ou=entapp

$ ldapmodify -p 3060 -h localhost -D "cn=orcladmin" -w passwd -f assigngrp.ldif
add uniqueMember:
 cn=emp1234,ou=people,ou=entapp
modifying entry cn=Administrators,ou=groups,ou=entapp
modify complete

Though the above example was derived from an Oracle Internet Directory (OID) environment, the problem and the solution are applicable to all environments running LDAP servers.

Read More

Session Sharing with another User on *NIX and Windows

Oracle Solaris

Since Solaris is not widely known for its graphical interface, let's just focus on sharing a terminal session in read-only mode with another user on the same system. Here is an example.

eg.,

% finger
Login       Name               TTY         Idle    When    Where
root     Super-User            pts/1            Sat 16:57  dhcp-amer-vpn-rmdc-a
sunperf         ???            pts/2          4 Sat 16:41  pitcher.sfbay.sun.com

In this example, two users root and sunperf are connected to the same system from two different terminals pts/1 and pts/2 respectively. If the root user wants to show something to sunperf user -- what s/he is doing in her/his terminal, for example, it can be accomplished with the following command.

script -a /dev/null | tee -a <target_terminal>

eg.,

# script -a /dev/null | tee -a /dev/pts/2
Script started, file is /dev/null
                                 # 
# uptime
  5:04pm  up 1 day(s),  2:56,  2 users,  load average: 0.81, 0.81, 0.81
# 
# isainfo -v
64-bit sparcv9 applications
        crc32c cbcond pause mont mpmul sha512 sha256 sha1 md5 camellia kasumi 
        des aes ima hpc vis3 fmaf asi_blk_init vis2 vis popc 
32-bit sparc applications
        crc32c cbcond pause mont mpmul sha512 sha256 sha1 md5 camellia kasumi 
        des aes ima hpc vis3 fmaf asi_blk_init vis2 vis popc v8plus div32 mul32 
# 
# exit
Script done, file is /dev/null

After the script .. | tee .. command, sunperf user should be able to see the root user's stdin and stdout contents in her/his own terminal until the script session exits in root user's terminal. Since this kind of sharing is based on capturing and redirecting the contents to the target terminal, the users on the receiving end won't be able to see whatever is being edited on initiators' terminal [using editors such as vi]. Also it is not possible to share the session with any connected user on the system unless the initiator has the necessary permissions and privileges.

The script utility records everything printed in a terminal session, while the tee utility replicates the contents of the screen capture on to the standard output of the target terimal. The tee utility does not buffer the output - so, the screen capture from the initiators' terminal appears almost right away in the target terminal.

Though I never tested, this technique may work on all *NIX and Linux flavors with little or no changes. Also there might be other ways to accomplish this.

[Thanks to Sujeet for sharing this tip]

Microsoft Windows

Most of the Windows users may rely on VNC services to share a desktop session. Another way to share the desktop session is to use the Remote Desktop Connection (RDC) client. Here are the steps.

• Connect to the target Windows system using Remote Desktop Connection client
  
• Launch Windows Task Manager
  
• Navigate to the "Users" tab
  
• Find the user session that you want to connect to and have full control over as the other user who is currently holding that session
  
• Select the user name in Windows Task Manager, right click and choose the option "Remote Control"
  
• A window pops up on the other user's session with the message "<USER> is requesting to control your session remotely. Do you accept the request?"

Once the other user says "Yes", you will be granted access to that session. Since then both users should be able to see the same screen and even control the session from their respective workstations.

Read More

Oracle E-Business Suite Tip : SQL Tracing

Issue:

Attempts to enable SQL tracing from concurrent request form fails with error:

Function not available to this responsibility. Change Responsibilities or contact your System Administrator

Resolution:

Switch responsibility to "System Administrator". Navigate to System -> Profiles, and query for "%Diagnostics% ("Utilities : Diagnostics")". Once found the profile, change its value to "Yes". Restart web browser and try enabling SQL trace again.

Read More

OBIEE 11g: Resolving Presentation Services Startup Failure

ISSUE:

Starting Presentation Services fail with the error:

[OBIPS] [ERROR:1] [] [saw.security.odbcuserpopulationimpl.getbisystemconnection] [ecid: ] [tid: ] Authentication Failure.
Odbc driver returned an error (SQLDriverConnectW).
State: 08004.  Code: 10018.  [NQODBC] [SQL_STATE: 08004] [nQSError: 10018] Access for the requested connection is refused.
[nQSError: 43113] Message returned from OBIS.
[nQSError: 43126] Authentication failed: invalid user/password. (08004)[[

Also connecting to the metadata repository (RPD) in online mode fails with similar error.

Looking through the BI server log, nqserver.log, you may find an error message similar to the following:

[OracleBIServerComponent] [ERROR:1] [] [] [ecid: 0001J1LfUetFCC3LVml3ic0000pp000000] [tid: 1] 
[13026] Error in getting roles from BI Security Service:  
'Error Message From BI Security Service: [nQSError: 46164] HTTP Server returned 404 (Not Found) for URL .' ^M

RESOLUTION:

• Connect to WebLogic Server (WLS) Console -> Deployments. Ensure that all deployed components are in 'Active' state.

• If any of the components is in 'Prepared' state, select that application and then click on "start servicing all requests"

• Restart BI Server and Presentation Services

In some cases, the following additional step might be needed to resolve the issue.

• Access the Enterprise Manager Fusion Middleware control: http://<host.domain>:port/em

• Navigate to Business Intelligence -> coreapplication

• 'Capacity Management' tab -> 'Scalability' sub-tab

• Click on 'Lock and Edit Configuration' button

• Enter the IP address in the 'Listen Address' field

• Click on 'Activate Changes' followed by 'Release Configuration' buttons
  

• Restart BI Server and Presentation Services

Also check these My Oracle Support (MOS) documents for more clues and information.

1387283.1 Authentication failed: invalid user/password
1251364.1 Error: "[nQSError: 10018] Access .. Refused. [nQSError: 43126] Authentication Failed .." when Installing OBIEE 11g
1410233.1.1 How To Bind Components / Ports To A Specific IP Address On Multiple Network Interface (NIC) Machines

Read More

Solaris Volume Manager (SVM) on Solaris 11

SVM is not installed on Solaris 11 by default.

# metadb
-bash: metadb: command not found

# /usr/sbin/metadb
-bash: /usr/sbin/metadb: No such file or directory

Install it using pkg utility.

# pkg info svm
pkg: info: no packages matching the following patterns you specified are
installed on the system.  Try specifying -r to query remotely:

        svm

# pkg info -r svm
          Name: storage/svm
       Summary: Solaris Volume Manager
   Description: Solaris Volume Manager commands
      Category: System/Core
         State: Not installed
     Publisher: solaris
       Version: 0.5.11
 Build Release: 5.11
        Branch: 0.175.0.0.0.2.1
Packaging Date: October 19, 2011 06:42:14 AM 
          Size: 3.48 MB
          FMRI: pkg://solaris/storage/svm@0.5.11,5.11-0.175.0.0.0.2.1:20111019T064214Z

# pkg install storage/svm
           Packages to install:   1
       Create boot environment:  No
Create backup boot environment: Yes
            Services to change:   1

DOWNLOAD                                  PKGS       FILES    XFER (MB)
Completed                                  1/1     104/104      1.6/1.6

PHASE                                        ACTIONS
Install Phase                                168/168 

PHASE                                          ITEMS
Package State Update Phase                       1/1 
Image State Update Phase                         2/2 

# which metadb
/usr/sbin/metadb

This time metadb may fail with a different error.

# metadb
metadb: <HOST>: /dev/md/admin: No such file or directory

Check if md.conf exists.

# ls -l  /kernel/drv/md.conf 
-rw-r--r--   1 root     sys          295 Apr 26 15:07 /kernel/drv/md.conf

Dynamically re-scan md.conf so the device tree gets updated.

# update_drv -f md

# ls -l  /dev/md/admin
lrwxrwxrwx   1 root root 31 Apr 20 10:12 /dev/md/admin -> ../../devices/pseudo/md@0:admin

# metadb
metadb: <HOST>: there are no existing databases

Now Solaris Volume Manager is ready to use.

eg.,
#  metadb -f -a c0t5000CCA00A5A7878d0s0

# metadb
        flags           first blk       block count
     a        u         16              8192          /dev/dsk/c0t5000CCA00A5A7878d0s0

Read More

Resolving "PLS-00201: identifier 'DBMS_SYSTEM.XXXX' must be declared" Error

Here is a failure sample.

SQL> set serveroutput on
SQL> alter package APPS.FND_TRACE compile body;

Warning: Package Body altered with compilation errors.

SQL> show errors
Errors for PACKAGE BODY APPS.FND_TRACE:

LINE/COL ERROR
-------- -----------------------------------------------------------------
235/6    PL/SQL: Statement ignored
235/6    PLS-00201: identifier 'DBMS_SYSTEM.SET_EV' must be declared
..

By default, DBMS_SYSTEM package is accessible only from SYS schema. Also there is no public synonym created for this package. So, the solution is to create the public synonym and grant "execute" privilege on DBMS_SYSTEM package to all database users or a specific user.

eg.,

SQL> CREATE PUBLIC SYNONYM dbms_system FOR dbms_system;

Synonym created.

SQL> GRANT EXECUTE ON dbms_system TO APPS;

Grant succeeded.

- OR -

SQL> GRANT EXECUTE ON dbms_system TO PUBLIC;

Grant succeeded.

SQL>  alter package APPS.FND_TRACE compile body;

Package body altered.

Note that merely granting execute privilege is not enough -- creating the public synonym is as important to resolve this issue.

Read More

Oracle RDBMS & Solaris : Few Random Tips (Feb 2012)

These tips are just some quick solutions or workarounds. Use these quickies at your own risk.

[#1] Oracle Data Pump

Q: How to exclude the table definition while importing a table using Oracle Data Pump import utility?

A: Use EXCLUDE=TABLE/TABLE option.

eg.,

impdp login/password DUMPFILE=<DUMP_FILENAME> LOGFILE=<LOGFILE_NAME> \
 DIRECTORY=<DB_DIR_NAME> TABLES=<TABLE_NAME> EXCLUDE=TABLE/TABLE

---

[#2] Workaround to ORA-01089: immediate shutdown in progress - no operations are permitted

When the database is in the middle of an instance shutdown, if another shutdown or startup was attempted, Oracle RDBMS may throw the above ORA-01089 error. The workaround is to force Oracle to start the database instance using startup force option. This option will shutdown the database instance (if running) using the abort command and then starts it up.

eg.,

SQL> STARTUP FORCE

---

[#3] Quick steps to upgrade the Oracle database from version 11.2.0.[1 or 2] to 11.2.0.3

Execute the following in the same sequence as sysdba.

startup upgrade
!cd $ORACLE_HOME/rdbms/admin
@utlu112i.sql  /* pre-upgrade information tool */
exec dbms_stats.gather_dictionary_stats (DEGREE => 64);
@catupgrd.sql  /* create/modify data dictionary tables */
@utlu112s   /* all components should be in VALID state */
shutdown immediate
startup
@catuppst.sql  /* upgrade actions that do not require DB in UPGRADE mode */
@utlrp.sql  /* recompile stored PL/SQL and Java code */
SELECT count(*) FROM dba_invalid_objects;  
                        /* verify that all packages and classes are valid */
exit

---

[#4] Q: Solaris: how to get rid of zombie processes?

A: Run the following with appropriate user privileges.

ps -eaf | grep defunct | grep -v grep | preap `awk '{ print $2 }'`

Alternative way: (not as good as the previous one - still may work as expected)

prstat -n 500 1 1 | grep zombie | preap `awk '{ print $1 }'`

---

[Added on 03/01/2012]

[#5] Solaris: Many TCP listen drops

eg.,

# netstat -sP tcp | grep tcpListenDrop
        tcpListenDrop       =2442553     tcpListenDropQ0     =     0

To alleviate numerous TCP listen drops, bump up the value for the tunable tcp_conn_req_max_q

# ndd -set /dev/tcp tcp_conn_req_max_q <value>

---

[Added on 03/02/2012]

[#6] Solaris ZFS: listing all properties and values for a zpool

Run: zfs get all <zpool_name> as any OS user

eg.,

% zpool list
NAME    SIZE  ALLOC   FREE    CAP  HEALTH  ALTROOT
rpool   276G   167G   109G    60%  ONLINE  -
spec    556G   168G   388G    30%  ONLINE  -

% zfs get all rpool
NAME   PROPERTY              VALUE                  SOURCE
rpool  type                  filesystem             -
rpool  creation              Fri May 27 17:06 2011  -
...
rpool  compressratio         1.00x                  -
rpool  mounted               yes                    -
rpool  quota                 none                   default
rpool  reservation           none                   default
rpool  recordsize            128K                   default
...
rpool  checksum              on                     default
rpool  compression           off                    default
...
rpool  logbias               latency                default
rpool  sync                  standard               default
rpool  rstchown              on                     default

---

[#7] Solaris: listing all ZFS tunables

Run: echo "::zfs_params" | mdb -k with root/super-user privileges

eg.,

# echo "::zfs_params" | mdb -k
arc_reduce_dnlc_percent = 0x3
zfs_arc_max = 0x10000000
zfs_arc_min = 0x10000000
arc_shrink_shift = 0x5
zfs_mdcomp_disable = 0x0
zfs_prefetch_disable = 0x0
..
..
zio_injection_enabled = 0x0
zvol_immediate_write_sz = 0x8000

Read More

Solaris Tip: How-To Identify Memory Mapped Files

A memory mapped (mmap'd) file is a shared memory object, or a file with some portion or the whole file was mapped to virtual memory segments in the address space of an OS process. Here is one way to figure out if a given object (file or shared memory object) was memory mapped in a process or not.

1. find the file system inode number of the object
2. look for that inode number in the address space of a given process

And here is an example. We are about to check a log file and a shared memory segment in a Siebel object manager's process address space.

# pfiles 8251
8251:   siebmtshmw /siebel/siebsrvr/admin/Siebel81.isve02.s
..
   1: S_IFREG mode:0744 dev:256,65539 ino:246660 uid:1234 gid:30 size:0
      O_WRONLY|O_APPEND|O_CREAT
      /siebel/siebsrvr/enterprises/Siebel81/isve02/log/StdErrOut/stderrout_8251_23311913.log
...
   9: S_IFREG mode:0700 dev:256,65539 ino:246640 uid:1234 gid:30 size:6889472
      O_RDWR|O_CREAT|O_EXCL
      /siebel/siebsrvr/admin/Siebel81.isve02.shm
..

# pmap -sx 8251 | grep 246660    
#               <== stderrout_8251_23311913.log file was not a memory mapped file

# pmap -sx 8251 | grep 246640
F6400000      64      64       -       -   8K r--s-  dev:256,65539 ino:246640   
F6410000     136     136       -       -    - r--s-  dev:256,65539 ino:246640
F6432000     128     128       -       -   8K r--s-  dev:256,65539 ino:246640
...
                <== Siebel81.isve02.shm was a memory mapped object

Read More

Unwanted Software Installers

After all these years of software evolution, it is odd to see not much improvement in the area of software installation. Customers do not seem to mind dealing with different, complex installers. Nevertheless this whole process can be simplified to save time, effort and energy.

In an ideal world, a software installer is supposed to have just one function - copying the software bits to a designated location and nothing else. However today we interact with different installers that does variety of things -- some install the pre-compiled binaries, few come in ready-to-extract zip archives, few others compile the binary on-the-fly and install the binary. Most of the enterprise software installers configure the software as part of the installation process where as few installers install the software and simply quit leaving the configuration step for the experts. Some of the installers hard-code the hostname, IP address, absolute paths of certain files etc., into some of the files on target system, which makes it hard to re-use the software home directory on a different server. Few installers do sensible job by not tying anything to the host system where the software is being installed.

Here is my partial wish list of features for a software installer. I think it is enough to make a point.

— Idempotent installations : install the software once, run anywhere. Customers should be able to move the resulting home directories from one host to any location on another host without worrying about the underlying changes to the location of the home directory, hostname, IP address etc., One example is the Oracle RDBMS installation. Once installed, the ORACLE_HOME can be zipped up, moved to another host, extracted and used right away. ORACLE_HOME usually contains the binaries. Installation specific configuration is stored outside of ORACLE_HOME. Optional Oracle Grid Control configuration appears to be saved under ORACLE_HOME, which is an aberration though it can be easily reconfigured once moved to another host.

— Simplicity : providing the entire directory structure in an extractable compressed archive file will remove one or more layers of dependency that the software installer has. For example, some of the installers require Java run-time to show the graphical interface for the installer. I recently encountered an installer executable that has private/unsupported symbols statically linked to it. When those private interfaces were removed in a later version of the operating environment, installer crashed and failed to make any progress. Had the software been provided in an extractable archive, software would have been readily available in the latter case. It appears that Oracle Corporation is moving in the right direction by releasing WebLogic 12c software as a zip file.

— De-couple software installation from configuration : there should be clear separation between the installation and configuration. Once the software is in place, relevant folks can always configure the software as directed and needed. The customer just needs a simple tool or script to configure the software.

        => Off-topic: providing a web interface is even better. It gives the flexibility to configure the software from anywhere in the same network.

— Contain everything in a single top-level directory : it makes patching easier even if the top-level directory was moved to a different location or host. No point in spreading the pieces of software into multiple locations anyway. Going back to the example of ORACLE_HOME, one shortcoming in Oracle RDBMS installation is that few directories/files such as oraInventory reside outside of ORACLE_HOME - so, when moving ORACLE_HOME to another host, it is necessary to move all relevant files that are outside of ORACLE_HOME as well for successful database software patching.

With careful planning/design, Release Engineering can be as creative and innovative as the rest of the teams in delivering a software product out of the door --- but I guess it is up to the customers to demand that attitude.

PS:
This blog post can be improved a lot. However since it is mostly about an opinion and a wish list, there is not much motivation to put more effort into it. And of course it is a generic discussion - nothing specific to a particular software or corporation.

Read More