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

Labels: how-to solaris mmap

Posted by: Giri Mandalika.    # 6:46 PM 0 comments.     Links to this post

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.

Labels: generic+discussion software installer

Posted by: Giri Mandalika.    # 11:43 PM 0 comments.     Links to this post

OATS is a suite of applications that can be used for performance and scalability testing, functional and regression testing. It is a thin client application that runs within a web browser - so, it is easy to use the tool from anywhere as long as the web server running on the host node is accessible. Hopefully the following tips and tricks will benefit some of the users of the Oracle Application Testing Suite.

Few technical details first - OATS is a 32-bit Java application that runs in a WebLogic container (WLS) with Oracle XE database being the backend store for test session data.

[Trick] Issue : OATS software fails to install on 64-bit Windows systems

Resolution:
Download and install 64-bit .NET framework manually before installing the OATS software. Look for .NET framework on Microsoft's downloads website.

[Trick] Issue : OATS software fails to install on systems with large number of [virtual] CPUs

Resolution:
On systems with many cores/vCPUs, Oracle database in general requires large amounts of memory to be configured for SGA - so, one solution would be to allocate as much memory as required. However Oracle XE limits the memory utilization within the database to 1 GB. Besides, XE uses only one CPU even if there are multiple CPUs available on a system. Hence one workaround is to limit the number of vCPUs that the system exposes during the installation of OATS software. The steps are shown below.

• Start button -> Run -> type "msconfig"
• Click on Boot tab -> Advanced Options
• Check "Number of processors" and set appropriate value (I believe we can go up to 16)
• Reboot Windows
• Uninstall failed OATS installation and try installing again
• Undo the above made changes after the successful installation of OATS
• Reboot Windows one final time

Thanks to my colleague Bao Doan for providing this workaround.

[Trick] Issue : During runtime, OATS drive the load and executes the test as expected but fails to collect runtime statistics

Resolution:
This is another limitation of Oracle XE database. Until 10g, XE limits the maximum amount of user data in the database to 4 GB. This limit was raised to 11 GB in release Oracle 11g XE. OATS 9.x releases bundle Oracle 10g XE. To take advantage of the larger limit for data, install Oracle 11g XE manually before installing OATS software. OATS installer gives the option to use an existing installation of Oracle XE. Besides, it is not possible to have multiple Oracle XE installations on a single box anyway (that's another XE limitation).

For existing installations, one workaround is to remove old and unwanted sessions to make room for new sessions in the database. Listed below are the steps.

• Connect to the Oracle Load Testing (OLT) tool
• Click on "Manage" top-level menu (upper right corner) -> Sessions
• Click on any unwanted session and press "Delete" button (I recommend deleting one session at a time)

[Trick] Issue : Under load, there are many network timeouts with ton of sockets in TIME_WAIT state on OATS agent systems including the OATS Controller node

Resolution:
Tune TCP/IP parameters on Windows as shown below.

• Launch Windows registry
• Navigate to HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\TcpIP\parameters
• Configure the following two parameters. If not found, create those parameters by selecting Edit -> New -> DWORD Value from the menu bar. Select "Decimal" under Base.
  TcpTimedWaitDelay : 30 [seconds]
  MaxUserPort : 65534
• Reboot Windows

Thanks to my colleagues Dino and Vishnu for sharing this workaround.

[Trick] Issue : OATS Controller does not show any graphs or analysis reports

Resolution:
Install Adobe Flash Plugin and try again.

[Trick] Issue : Under load, OATS Controller stops collecting runtime statistics at some random point

Resolution:
Check Oracle database alert log for some clue(s). If there is an error message such as "ORA-12516: TNS:listener could not find available handler with matching protocol stack", connect to the database, query v$resource_limit view and compare the values reported under CURRENT_UTILIZATION and MAX_UTILIZATION for the resource "processes". If the current utilization is pretty close to the configured maximum value, raise the value for processes parameter in [S]PFILE.

[Tip] Balancing the load among multiple OATS agent systems

One simple way is to create a VU Agent System Group based on the available agent systems. Steps listed below.

• Connect to the Oracle Load Testing (OLT) tool
• Click on "Manage" top-level menu (upper right corner) -> Systems
• Click on "VU Agent System Group" in the left hand side
• On the right hand side, click on "New" option
• Select all the agent systems that you want to be part of the "VU Agent System Group"
• Finally name the newly created system group and save

Note that it is not possible to attach weights to the agent systems - so, it is suggested to have agent systems with similar hardware configurations in the VU Agent System Group.

[Tip] Balancing the load among multiple web servers using OATS Controller

If there are multiple web server instances running in a enterprise application deployment; and OATS software is being used to test the performance and scalability of the application, parameterizing the web server hostname and port number in OATS test script will take care of the web server load balancing problem. Of course there are many alternatives to this approach such as using a hardware load balancer, using web server Reverse Proxy etc.,

[Added on 01/19/2012]

[Tip] How-To check the available space in USERS tablespace?

Run the following on OATS Controller node:

Start -> All Programs -> Oracle Database XX Express Edition -> Run SQL Command Line

SQL> connect / as sysdba

SQL> SELECT /* + RULE */  df.tablespace_name "Tablespace",
       df.bytes / (1024 * 1024) "Size (MB)",
       SUM(fs.bytes) / (1024 * 1024) "Free (MB)",
       Nvl(Round(SUM(fs.bytes) * 100 / df.bytes),1) "% Free",
       Round((df.bytes - SUM(fs.bytes)) * 100 / df.bytes) "% Used"
  FROM dba_free_space fs,
       (SELECT tablespace_name,SUM(bytes) bytes
          FROM dba_data_files
         GROUP BY tablespace_name) df
 WHERE fs.tablespace_name (+)  = df.tablespace_name
 GROUP BY df.tablespace_name,df.bytes
UNION ALL
SELECT /* + RULE */ df.tablespace_name tspace,
       fs.bytes / (1024 * 1024),
       SUM(df.bytes_free) / (1024 * 1024),
       Nvl(Round((SUM(fs.bytes) - df.bytes_used) * 100 / fs.bytes), 1),
       Round((SUM(fs.bytes) - df.bytes_free) * 100 / fs.bytes)
  FROM dba_temp_files fs,
       (SELECT tablespace_name,bytes_free,bytes_used
          FROM v$temp_space_header
         GROUP BY tablespace_name,bytes_free,bytes_used) df
 WHERE fs.tablespace_name (+)  = df.tablespace_name
 GROUP BY df.tablespace_name,fs.bytes,df.bytes_free,df.bytes_used
 ORDER BY 4 DESC;

Copy/paste the above SQL code in a text file with sql extension and execute that SQL statement by calling the SQL script from SQL> command prompt. eg., assuming the above code was saved in a plain text file called chktblspcusg.sql under C:\ drive, execute the SQL script as shown below:

SQL> @C:\chktblspcusg.sql

See Also:

• Oracle Application Testing Suite
• Oracle Database XE Licensing Restrictions
• Check the documentation that is available as part of OATS installation

Labels: OATS Oracle

Posted by: Giri Mandalika.    # 3:14 PM 0 comments.     Links to this post

Symptom:

System log shows a bunch of RPC timed out messages such as the following:

Dec 13 09:23:23 gil08 last message repeated 1 time
Dec 13 09:29:14 gil08 statd[19858]: [ID 766906 daemon.warning] statd: cannot talk to statd at ssc23, RPC: Timed out(5)
Dec 13 09:35:05 gil08 last message repeated 1 time
Dec 13 09:40:56 gil08 statd[19858]: [ID 766906 daemon.warning] statd: cannot talk to statd at ssc23, RPC: Timed out(5)
..

Those messages are the result of an apparent communication failure between the status daemons (statd) of both local and remote hosts using RPC calls.

Workaround/Solution:

If the target_host is reachable, execute the following to stop the system from generating those warning messages --- stop the network status monitor, remove the target host entry from /var/statmon/sm.bak file and start the network status monitor process. Removing the target host entry from sm.bak file keeps that machine from being aware that it may have to participate in locking recovery.

eg.,

# ps -eaf | fgrep statd 
  daemon 14304 19622   0 09:47:16 ?           0:00 /usr/lib/nfs/statd
    root 14314 14297   0 09:48:03 pts/15      0:00 fgrep statd

# svcs -a | grep "nfs/status"
online          9:52:41 svc:/network/nfs/status:default

# svcadm -v disable nfs/status
svc:/network/nfs/status:default disabled.

# ls /var/statmon/sm.bak
ssc23

# rm /var/statmon/sm.bak/ssc23

# svcadm -v enable nfs/status
svc:/network/nfs/status:default enabled.

Labels: oracle solaris rpc statd RPC troubleshooting

Posted by: Giri Mandalika.    # 7:45 PM 0 comments.     Links to this post

Symptom:

A newly installed Siebel application server fails to start despite successful ODBC connectivity to the database. SRProc process logs ODBC error messages similar to the following:

Message: GEN-13,
 Additional Message: dict-ERR-1109: 
       Unable to read value from export file (Data length (32) > Column definition (3)).

Message: GEN-13,
 Additional Message: dict-ERR-1107: Unable to read row 0 from export file (UTLDataValRead pBuf, col 4 ).

GenericLog  GenericError  1     0002157..  11-11-18 13:28  Message: Generated SQL statement:,
 Additional Message: SQLFetch:
   SELECT RDOBJ.DOCK_ID, RDOBJ.RELATED_DOCK_ID, RDOBJ.SQL_STATEMENT, RDOBJ.CHECK_VISIBILITY,
          'N', RDOBJ.COMMENTS, RDOBJ.ACTIVE, RDOBJ.SEQUENCE, RDOBJ.VIS_STRENGTH,
          RDOBJ.REL_VIS_STRENGTH, RDOBJ.VIS_EVT_COLS
     FROM ORAPERF.S_DOCK_REL_DOBJ RDOBJ, ORAPERF.S_DOCK_OBJECT DOBJ
    WHERE RDOBJ.REPOSITORY_ID = (SELECT ROW_ID FROM ORAPERF.S_REPOSITORY WHERE NAME = ?)
      AND DOBJ.ROW_ID = RDOBJ.DOCK_ID
      AND (DOBJ.INACTIVE_FLG = 'N' OR DOBJ.INACTIVE_FLG IS NULL)
      AND (RDOBJ.INACTIVE_FLG = 'N' OR RDOBJ.INACTIVE_FLG IS NULL)

Message: Error: An ODBC error occurred,
 Additional Message: Function: DICGetRDObjects; ODBC operation: SQLFetch

Message: GEN-13,
 Additional Message: dict-ERR-1109: Unable to read value from export file (UTLCompressFRead (fseek)).

Message: GEN-13,
 Additional Message: dict-ERR-1107: Unable to read row 0 from export file (UTLDataValRead pBuf, col 0 ).

Message: GEN-10,
 Additional Message: Calling Function: DICLoadDObjectInfo; Called Function: Calling DICGetRDObjects

Message: GEN-10,
 Additional Message: Calling Function: DICLoadDict; Called Function: DICLoadDObjectInfo

GenericError
(srpdb.cpp (860) err=3006 sys=2) SBL-GEN-03006: Error calling function: DICFindTable m_pReqTbl
(srpsmech.cpp (74) err=3006 sys=0) SBL-GEN-03006: Error calling function: DICFindTable m_pReqTbl
(srpmtsrv.cpp (107) err=3006 sys=0) SBL-GEN-03006: Error calling function: DICFindTable m_pReqTbl
(smimtsrv.cpp (1203) err=3006 sys=0) SBL-GEN-03006: Error calling function: DICFindTable m_pReqTbl
SmiLayerLog Error       Terminate process due to unrecoverable error: 3006. (Main Thread)

An inconsistent or corrupted dictionary file "diccache.dat" is likely the cause.

Solution:

• Stop the application server and manually kill the remaining Siebel application specific processes

  eg.,

  stop_server all
  
  pkill siebmtsh
  pkill siebproc
  ..
  

• Remove $SIEBEL_HOME/bin/diccache.dat file. It will be re-generated during the application server startup

• Start the application server

  start_server all
  

Posted by: Giri Mandalika.    # 3:37 PM 0 comments.     Links to this post