Friday, January 31, 2014

DataPump impdp Raises Error ORA-942 When Creating Index

While importing a database using impdb you can get ORA-00942 error during an index creation. As you know ORA-00942 means table or view does not exist. So when you see this error in an index creation, Oracle actually says that : I cant create the index because the underlying table is not there, even if it is there I can not the see the table because you dont have necessary privileges :)
That is, this error can be generated if the calling schema/user doesn't have the required privileges for the action, as well
Also note that even if we do the import with sys user , we can still get this error in an import process..
The reason behind that is the design of impdb.. So it is expected.. That is ; When impdp creates indexes, it connects to the schema owner before issuing the CREATE INDEX command. So if the schema owner does not have the privilege to create an index on a table, the error ORA-00942 is generated.

For the solution, grant necessary privileges to the user which is the owner of the index..
Note that , you can always create an index later.. So if you have done your import and if you dont want to repeat it, then parse the logfile, find the problematic index creations and create them manually.

EBS-- impdb ORA-39083, ignorable APPS trigger creation error

During the import process of an EBS(11i -- database using impdb , you can get ORA-39083 and ORA-25001 errors..
ORA-39083: Object type TRIGGER failed to create with error:
ORA-25001: cannot create this trigger type on views.

Note that the triggers below are obsolete in 11.5.1 or higher:

The related base tables are still exists but they are not used anymore. They are replaced with TCA architecture (HZ_ tables).
So if the trigger that this error is produced for, is one of the above , then the error is ignorable.

This is well documented in MOS DOC: 1170463.1

RDBMS --expdp ORA-39095: Dump file space has been exhausted

When exporting the whole database using datapump, ORA-39095 can be encountered. It seems it is caused by an unpublished bug 7486649 and fixed in version 11.2..
The reason behind the error seems to be the FILESIZE parameter used with expdp command..
The workaround is to increase the size used in FILESIZE parameter or to not to use the FILESIZE parameter at all..

I have faced this problem with a EBS database. 
I checked Oracle Support. Related Oracle MOS documents say using the dynamic format can prevent this error to arise, but I already tried that and it was not the solution for me.

The solution I applied was increasing the FILESIZE parameter. By increasing the FILESIZE parameter you will have less number of dump file pieces and you wont get ORA-39095 while exporting your big size database. Because it seems sometimes datapump throws this error even if you set your dumpfile parameter to support several number of dump file pieces (like 99 -> with dumpfile=fullexp%U.dmp)

Thursday, January 30, 2014

EBS -- Datapump expdp error ORA-33272 , ZPB, DBMS_AW, dba_aws

While exporting an EBS 11i database using datapump , you can get ORA-33272 error on Analytical Workspaces..
This errors are mostly due to the versions of aw objects..
Suppose you have upgraded your Ebs database from 9i to 10g and the AW objects belong to 9i version of Oracle Database have remained in your database..  In such a situation, you can encounter ORA-33272 error on your future data pump exports..

An example for these kind of errors are below;

ORA-39127: unexpected error from call to export_string :=SYS.DBMS_AW_EXP.schema_info_exp('ZPB',1,1,'',newblock)
ORA-33272: Analytic workspace ZPB.ZPBCODE cannot be opened.
ORA-00600: internal error code, arguments: [xschkmk:!dict], [], [], [], [], [], [], []
ORA-06512: at "SYS.DBMS_AW", line 67
ORA-06512: at "SYS.DBMS_AW", line 131
ORA-06512: at "SYS.DBMS_AW", line 792
ORA-06512: at "SYS.DBMS_AW", line 1142
ORA-06512: at "SYS.DBMS_AW_EXP", line 517
ORA-06512: at line 1
ORA-06512: at "SYS.DB.

For the solution, use following query to list the version for the object in question;

select * from all_aws where owner='ZPB';  (ZPB is the schema name of failing objects in the example above)

SQL> select * from dba_aws;

Example output:
---------- ---------- -------------------- ---- ---------- ----------- ------
ZPB 1000 ZPBCODE 9.1 292 2
SYS 129 AWREPORT 10.2 9 2
SYS 128 AWXML 10.2 64 2
ZPB 1001 ZPBANNOT 9.1 7 1
ZPB 1002 ZPBDATA 9.1 7 1
SYS 125 AWMD 10.2 363 2
SYS 124 EXPRESS 10.2 46 2
SYS 126 AWCREATE 10.2 21 2
SYS 127 AWCREATE10G 10.2 9 2

Analyze the output of the query and if you have older versions, delete them..
For the example above: the ZPB Aws are 9i and they are need to be deleted.

You can use dbms_aw.execute to do that..

exec dbms_aw.execute('aw delete zpb.zpbcode');
exec dbms_aw.execute('aw delete zpb.zpbannot');
exec dbms_aw.execute('aw delete zpb.zpbdata');

Log in AS ZPB
SQL> set serverout on
SQL> call dbms_aw.execute('AW DELETE ZPBCODE');
SQL> call dbms_aw.execute('AW DELETE ZPBANNOT');
SQL> call dbms_aw.execute('AW DELETE ZPBDATA');
If you receive errors deleting the AWs, drop the associated AW$ tables :
SQL> drop table AW$ZPBCODE;
SQL> drop table AW$ZPBANNOT;
SQL> drop table AW$ZPBDATA;

select * from dba_aws;

SYS 125 AWREPORT 9.1 9 2
SYS 124 AWXML 9.1 50 2
SYS 123 AWCREATE10G 9.1 9 2
SYS 122 AWCREATE 9.1 18 2
SYS 121 AWMD 9.1 205 2
SYS 120 EXPRESS 9.1 38 2

Also if you have all 9i 's in awr$ table, then you may be need to -> Remove OLAP, Clean up Left Over Objects And Add It back (Doc ID 1560841.1)

Is my index/index (sub)partition unusable? N/A state indexes, Exadata

When you work with Exadata , we usually compress the data /the tables.. One of the method to compress the data in the tables, which the compress atrribute set on, is moving the table with "alter table move command" .. Note that, Exadata is a preferred system for large databases in which the tables are partitioned even subpartitioned. So in that case; you need move these partitions and subpartitions, too..

When the tables/partitions/subpartitions are moved, the indexes become invalid, as moving is a physical operation which causes the rowid stored in related indexes change..
So you need to rebuild the indexes , index partition and index partitions, accordingly...

To find the indexes those need to be rebuilt, you can use following queries.. The queries below are actually dynamic sqls which will produce the rebuild commands directly..

Sub partition rebuild:
select 'alter index '||index_owner||'.'||index_name||' rebuild subpartition '||subpartition_name||' parallel 8;' from dbA_ind_subpartitions where status not in ('USABLE','VALID')

Partition rebuild:
select 'alter index '||index_owner||'.'||index_name||' rebuild partition '||partition_name||' parallel 8;' from dbA_ind_partitions  where status not in ('USABLE','VALID')

Index rebuild:
select 'alter index '||owner||'.'||index_name||' rebuild parallel 8;' from dbA_indexes  where status not in ('USABLE','VALID')

The thing that me writing this post is the index status N/A.
There are indexes in N/A status and  N/A status does not mean that the index must be rebuilt..
You have to analyze this N/A status in levels. I mean, lets suppose dba_indexes = LEVEL1, dba_partitions=LEVEL2, dba_subpartitions=LEVEL3

So if you see a N/A state index in dba_indexes, check dba_ind_partitions..
And if you see a N/A state index partition in dba_ind_partitions check dba_ind_subpartitions :)
so check LEVEL1 -> LEVEL 2 -> LEVEL 3.
In brief, you dont rebuild every N/A state index you see :)

tnsing TNS-03507 Even Though SQL*Net Succeeds

This error can be encountered using tnsping binary stored in 8.0.6 Oracle Home;
Here is was happens,
[applexadata@ermanapp ~]$ tnsping EXADEV_806_BALANCE
TNS Ping Utility for Linux: Version - Production on 24-JAN-2014 16:24:39
(c) Copyright 1997 Oracle Corporation. All rights reserved.
TNS-03507: Failure looking for ADDRESS keyword

As you see above , we cant ping the defined listener through EXADEV_806_BALANCE using tnsping
But, when we use sqlplus stored in same 8.0.6 Oracle home to connect to the same database through the same tns entry, we can connect successfully.
sqlplus apps/apps@EXADEV_806_BALANCE;

SQL*Plus: Release - Production on Fri Jan 24 16:24:40 2014
(c) Copyright 1999 Oracle Corporation. All rights reserved.
Connected to:
Oracle Database 11g Enterprise Edition Release - 64bit Production
With the Partitioning, Real Application Clusters, Automatic Storage Management, OLAP,
Data Mining and Real Application Testing options
SQL>  --> connected successfully.

So this is the problem of tnsping utility stored in 8.0.6 Oracle Home. Maybe the reason is its low software version.. It seems it does not like the format of the tns entry EXADEV_806_BALANCE generated by autoconfig..

Ignore these tnsping error if your application works properly.. Does you EBS components connect to the database using the BALANCE tns properly? Because, this seems to be important.

Oracle describes this as follows :)

The tnsping/tnsping80 utility is not designed in such a way that it will parse a Load Balancing tnsnames.ora file. Since tnsping is a test utility and SQL*Plus connections work, there is no need for any modification in the tnsnames parameter file.

EBS on Oracle RAC / Exadata LOAD BALANCE connections

As you know, techonology used in EBS are actually a collection.. We have Forms, Reports, Concurrent Managers, Apache/Web Server, External Reporting tools etc.
These different technologies --components are connecting to the EBS databases seperately.
These situation brings us an opportunity to isolate database nodes according to the Application Components.
That is, if we want, we can configure these components(Apache, Conc Managers, Forms, Reports and or external tools) to use only Database Node 1 for its sessions or we can configure Apache to use Database Node 1 and Node 2 load balanced..

In a Rac Environment we usually prefer to use a load balanced connection style for EBS Application Tier components. Of course, sometimes in some environments; we feel the need to isolate database nodes according to the components, workload and etc..
In short, to make the long story short, In this post I will describe how to make these configurations in EBS.

Briefly, we have a file called context file in EBS and this file feeds the Autoconfig.. That is, Autoconfig reads this file to make the necessary configurations in EBS components.
So to able to make the configurations described above, we need to change the following lines in context file and run autoconfig.

<TWO_TASK oa_var="s_tools_twotask" osd="unix">EXADEV</TWO_TASK>
<TOOLS_DB_TWOTASK oa_var="s_tools_db_twotask">EXADEV_806_BALANCE</TOOLS_DB_TWOTASK>
<CP_TWOTASK oa_var="s_cp_twotask">EXADEV</CP_TWOTASK>
<TWO_TASK oa_var="s_weboh_twotask" osd="unix">EXADEV</TWO_TASK>
<WEBOH_DB_TWOTASK oa_var="s_weboh_db_twotask">EXADEV_806_BALANCE</WEBOH_DB_TWOTASK>

We change the values stored in xml tags according to tnsnames.ora stored in relevant Oracle Homes ( I mean IAS ORACLE_HOME and 8.0.6 ORACLE_HOME for 11i and 10.1.2 ORACLE_HOME and 10.1.3 ORACLE_HOME for R12)

Normally, when we have 2 nodes database in EBS we have the following tns entries in our Application tier's ORACLE HOMES ..
Lets suppose our database Service Name is EXADEV and also suppose we dont have any scan listeners ( as you know they are not supported in EBS 11i)

So by looking to the tns entries above, we can change the context file and run the autoconfig to make the configurations according to our need..
For example;
According to the tns entries above, If we want an application tier component to connect to our cluster in a load balanced manner, we use EXADEV_BALANCE tns entry or If we have a compontent to use only the Database Node 2, we use EXADEV2 tns_entry and so on.

Hope this helps..

Of couse, we have parallel concurrent processing and Application node load balancing but these subject will be described in forthcoming blog posts of mine..

Wednesday, January 29, 2014

DBMS_STATS error gathering fixed object stats ORA-20000 ORA-6512 insufficient privileges

While gathering fixed objects status using DBMS_STATS.GATHER_FIXED_OBJECTS_STATS('ALL'), you can
encounter the following error;
ORA-20000 ORA-6512.. 
Error stack can be similar to following;
ORA-20000: Insufficient privileges to analyze an object in Fixed Ob
ORA-06512: at "SYS.DBMS_STATS"
ORA-06512: at "SYS.DBMS_STATS"
ORA-06512: at "SYS.DBMS_STATS"
ORA-06512: at line 1

Anyways, these errors are due to a error syntax.. 
Some Oracle documents say gather_fixed_objects_stats should be executed with 'ALL'  input,
but it seem they are wrong.. 
For example: E-Business Suite on Exadata Oracle Maximum Availability Architecture White Paper,October 2012
Because, DBMS_STATS.GATHER_FIXED_OBJECTS_STATS can take a parameter but it has
 to be an existingstatistics table created with DBMS_STATS.CREATE_STAT_TABLE.
(Reference Mos Document: 392364.1 )

For the solution; drop the 'ALL' argument and run the proc as follows;

Tuesday, January 21, 2014

Linux-- cron.daily hangs, mcelog hangs, Xen kernel, XenStoreD, /sys/hypervisor/uuid

In one of our client's production environments , we have faced an incident.
After startup (approximately 1 day after startup), the Linux server, which was a Redhat Linux 5.4 64 bit (with 2.6.18-164.el5xen  kernel) started to experience a high load.

Okay, I will keep it short;
The reason behing the problem was that the machine had 6 gig memory and it was all being used.  Actually this was due to a high process count on the Linux.. Because, when I checked processes with ps ; I saw that there were a lot of crond processes, the cron was trying to execute what was scheduled to run hourly (cron.hourly).. I didnt go in to details and analyze it further to prove that was the cause , because it was obviously the cause :)
Going down from the parent to child I saw that cron was trying to execute mcelog script, and the script was looping / hanging..
Going down from the parent to child processes again ; I saw that the script was trying to do the following;
cat /sys/hypervisor/uuid` != "00000000-0000-0000-0000-000000000000"
The cat command could not be finished. The relevant process was in D state, in other words waiting for I/O.

Then, I checked the mcelog dailiy script , it was like below;

if [ -e /proc/xen ] && [ `cat /sys/hypervisor/uuid` != "00000000-0000-0000-0000-000000000000" ]; 
# this is a PV Xen guest. Do not run mcelog.
 exit 1;
/usr/sbin/mcelog --ignorenodev --filter >> /var/log/mcelog

So, The crond was running mcelog.cron which is supplied by the mcelog package..
The kernel was a xen kernel, and xen related software packages was running on it..
Note that, there was no need for a Xen kernel to be used in this server, but it s another story :)

Brief explanatin about the Xen /hypervisor ,  as Xen is a native (bare-metal) hypervisor providing services that allow multiple computer operating systems to execute on the samecomputer hardware concurrently.


I found a similar bug in Suse, not Redhat Linux but it seems we were hitting something like the following,

patches.xen/499-sysfs-uuid-hang.patch: Avoid kernel hang reading
+ /sys/hypervisor/uuid if xenstore is not available.
Also look at the following;

Also this is from Redhat 5:

Bug 225203 Reading /sys/hypervisor/uuid in Dom0 hangs if XenStoreD isn't running
Description of problem: Attempting to read from /sys/hypervisor/uuid in Dom0 will hang indefinitely if XenStored hasn't been launched. Version-Release number of selected component (if applicable): uname -r 2.6.18-1.2747.el5xen

In conclusion;

we can say that, the mcelog was trying to gather some vm related information(vm domain based information) from XenStore, using Xenstore daemon, and Xenstore daemon was not running , we were hitting a bug in Redhat 5 caused  by the bug described above, as the /sys/hypervisor/uuid path relies on Xenstore to query the domain's UUID (handle) .

Possible solutions are:
  • Editing the mcelog.cron script, commenting all of the lines and putting and exit 0 at the end of the file..
  • Checking  with kill -0 `cat /var/run/`, before reading the /sys/hypervisor/uuid
  • The code in the script can change something like the following; ( I didnt test it)

           if [ -e /proc/xen/capabilities ] 
           # xen 
           grep control_d /proc/xen/capabilities >& /dev/null if [ $? -ne 0 ] 
           # domU -- do not run on xen PV guest 
           return 1; 

Monday, January 20, 2014

EBS-- Authorization Required , mod_plsql, fil, custom_auth

This problem arises in EBS, when the mod_plsql is in use. In my case, the victim was an Oracle Form that was trying to use mod_plsql to display some data..  (Note that , you can encounter this error in Apex, or in any Oracle Product that is using Oracle Application Server and mod_plsql..

The problematic web page was producing the following error in a little windows message box , which made me think it is a client error, in first sight.

Authorization Required

The error is similar to the following;
This server could not verify that you are authorized to access the document
requested. Either you supplied the wrong credentials (e.g., bad password), or
your browser doesn't understand how to supply the credentials required.

In addition , the Apache error log was displaying the following error.
HTTP-401 Authorization Required' error message

After some analysis, it could be figured out that, the problem was caused in the mod_plsql layer..
So lets describe what are the configuration file that is used by the mod_plsql to operate..

The file is where the DADs (Database Access Descriptors) are defined.. Using DADs, PL/SQL gateway(mod_plsql) can connect to the relevant databases..
Mod_plsql, in other words PLSQL gateway; is a part of Oracle HTTP Server (in this case Apache). Using mod_plsql ; a URL sent by a browser is interpreted, the associated plsql programs are called and the response is returned to calling browser..

Following describes the flow in mod plsql processing;
  • The Oracle HTTP Server receives a request containing a virtual path, which is configured to be serviced by mod_plsql.
  • The Oracle HTTP Server routes the request to mod_plsql.
  • By using the configuration information stored in your DAD, mod_plsql connects to the database. The request is forwarded by mod_plsql to the Oracle Database.
  • mod_plsql prepares the call parameters, and invokes the PL/SQL procedure in the application.
  • The PL/SQL procedure generates an HTML page using data and the PL/SQL Web Toolkit accessed from the database.
  • The response is returned to mod_plsql.
  • The Oracle HTTP Server sends the response to the client browser.
Okay, lets come back to our subject.. file..

Here is an example for file;

defaultDAD      = ERM
administrators  = system
adminPath       = /admin_/
custom_auth     = CUSTOM
connect_string  = PROD,REQUEST_PROTOCOL=http,SERVER_PORT=8000,
input_filtering = Yes
connect_string  = ERM
password        = erman
username        = APPS
default_page    =
document_table  = APPS.fnd_lobs_document
document_path   = docs
document_proc   =
upload_as_long_raw =
upload_as_blob  = *
reuse           = Yes
connmax         = 10
pathalias       = fndgfm
pathaliasproc   = fnd_gfm.dispatch
enablesso       = No,REQUEST_PROTOCOL=http,SERVER_PORT=8000,
input_filtering = Yes

Anyways, I will keep it short. 
Pay attention to the custom_auth parameter above, because that parameter is the cause of the error. This parameter is used  for custom authentication and is an old configuration used in ias and it causes problems with later mod_plsql applications.
For the solution, comment out the custom_auth parameter by putting a ";" character at the beginning of the line and bounce Oracle HTTP Server.

Wednesday, January 15, 2014

RDBMS-- Lock Compatability Matrix for Oracle Database

I thought that "Lock Compatability Matrix for Oracle Database" should be present in this blog.
Following table compares the locking operations, relevant lock modes and describes whether they are permitted.  

* Waits if another transaction has a lock.

If you interpret these tables; we can see that;
  • You can do a select in every time, selects are independent.. You dont have to wait for anything.
  • To able to do an insert , there should not be a Share or Shared Row Exclusive or Exclusive lock on the table. You dont have to wait for another transaction..
  • To able to do an update, there should not be a Share or a Shared Row Exclusive or an Exclusive lock on the table.You have to waits if another transaction has a lock
  • To able to do an delete, there should not be a Share or a Shared Row Exclusive or an Exclusive lock on the table.You have to waits if another transaction has a lock
  • To able to do a select for update, there should not be an Exclusive lock on the table, and you have to wait if someone has a lock that prevents you.You have to waits if another transaction has a lock
Another representation can be the following;

If you interpret these tables; we can see that;

For example, you cant have a share on a resource , if there is an exclusive lock on it, and iff you have a share lock on a resource, nobody can put an exclusive lock on that resource, and so on..

Glassfish -- Script to detect the failure and start the server again

If your glassfish server becomes terminated somehow, then you can use following scripts to start it again. This scripts will be executed by cron and will detect the failure and start the glassfish again..
I prepared them for Linux and Solaris..
Note that, this method can be applied to any other processes.. We simple grep the process from the output of ps -ef, and start it if necessary.


*/5 * * * * /root/erman_install/glassfish_restart_script ##  will be executed in every 5 mins

ps -ef |grep glassfish.jar |grep -v grep
if [ "$?" == "1" ]
echo "Glassfish process not found"
echo "Starting Glassfish"
sh /root/erman_install/glassfish3/glassfish/bin/asadmin start-domain  >/dev/null 2>&1
echo "Found Glassfish process. Dont need to start Glassfish"

Setup : Create the script , give the execute rights to relevant user , and make the cron entry.



0,5,10,15,20,25,30,35,40,45,50,55 * * * *  /erman_install/glassfish_restart_script   ##if your cron is not so old, you can use */5 too.. This will be executed in every 5 mins, too.

ps -ef |grep "/usr/java/bin/java -Dcom.sun.aas.instanceRoot=/erman_install/GlassFishESBv22/glas" |grep -v grep
if [ "$?" == "1" ]
echo "Glassfish process not found"
echo "Starting Glassfish"
sh /erman_install/GlassFishESBv22/glassfish/bin/asadmin start-domain  >/dev/null 2>&1
echo "Found Glassfish process. Dont need to start Glassfish"

Setup : Create the script , give the execute rights to relevant user , and make the cron entry.

Thursday, January 9, 2014

Linux & Oracle -- Using strace, pstack and /proc filesystem for analyzing processes.

This post will be about Linux process runtime analysis.
I choosed the DBWR process to analyze. I wonder what DBWR is doing in a certain time interval..
Is it working actually? Does it write or read? Which  file or files DBWR is working on?

I will try answer those question by making analysis on OS layer.
I will use top, stace, pstack and /proc filesystem to do that.

Lets start;
Here , we see DBWR is doing I/O. We use the top command  and see the DBWR process is in "D" state.

16748 dbexa     15   0 1212m 829m 787m D  0.3  4.6   0:24.39 ora_dbw0_EXADEV

We obtain the process id of dbwr from above. The process id of the dbwr process im question is 16748.

So lets attach that process using pstack and see what is in its execution stack..

we issue the command; pstack 16748

#0  0x00000034ce80bd05 in __pwrite_nocancel () from /lib64/tls/
#1  0x0000000003c147f9 in skgfqio ()
#2  0x000000000090af6e in ksfdgo ()
#3  0x0000000000907b4f in ksfdaio1 ()
#4  0x000000000090564a in ksfdaio ()
#5  0x00000000027292ae in kcfdowr ()
#6  0x00000000010e99ea in kcbbdrv ()
#7  0x00000000008270f1 in ksbabs ()
#8  0x000000000082f658 in ksbrdp ()
#9  0x0000000002f9b5e8 in opirip ()
#10 0x000000000134375e in opidrv ()
#11 0x0000000001ece70e in sou2o ()
#12 0x0000000000713b85 in opimai_real ()
#13 0x0000000000713a3c in main ()

So when we look above, we can see  __pwrite_nocancel from the library. So its coming from the threads library..
__pwrite_nocancel is in the top of the stack, so this function is the last thing that have been executed by this process. This process seems to be doing I/O, actually write I/O's..

Lets trace the system calls of the DBWR process using strace, to get a better runtime data for better understanding the I/O that the process does.

To do this, we issue the command strace -p 16748   (16748 is the process id of dbwr)

pwrite(116, "\2\242\0\0\244O\5\0303!\"\204\211\5\33\4\220\t\0\0;\0\31"..., 8192, 2851373056) = 8192
times(NULL)                             = 456992346
times(NULL)                             = 456992346
pwrite(116, "\2\242\0\0\252O\5\0303!\"\204\211\5\33\4\25\204\0\0;\0"..., 8192, 2851422208) = 8192
times(NULL)                             = 456992346
times(NULL)                             = 456992346
pwrite(116, "\2\242\0\0\254O\5\0303!\"\204\211\5\17\4\204m\0\0;\0\31"..., 8192, 2851438592) = 8192
times(NULL)                             = 456992346
times(NULL)                             = 456992347
pwrite(116, "\2\242\0\0\256O\5\0302!\"\204\211\5\33\4:X\0\0;\0\31\0"..., 8192, 2851454976) = 8192
times(NULL)                             = 456992347
times(NULL)                             = 456992347
pwrite(116, "\2\242\0\0\260O\5\0302!\"\204\211\5\33\4/\266\0\0;\0\31"..., 8192, 2851471360) = 8192
times(NULL)                             = 456992348
times(NULL)                             = 456992348
pwrite(116, "\2\242\0\0\262O\5\0302!\"\204\211\5\33\4\332\10\0\0;\0"..., 8192, 2851487744) = 8192
times(NULL)                             = 456992348
times(NULL)                             = 456992348
pwrite(116, "\2\242\0\0\264O\5\0302!\"\204\211\5\33\4\20-\0\0;\0\31"..., 8192, 2851504128) = 8192
times(NULL)                             = 456992351
times(NULL)                             = 456992351
pwrite(116, "\2\242\0\0\270O\5\0302!\"\204\211\5\33\4\316\311\0\0;\0"..., 8192, 2851536896) = 8192
Okay, we see a lot of pwrites and times sys calls.. pwrite syscalls we see here, are consistent with what we saw in the pstack (__pwrite_nocancel)

Lets look what this system calls actually are;

times(): returns the number of clock ticks that have elapsed since an arbitrary point in the past.
times() syscall is not so relevant with our approach..

pwrite() : write to a file descriptor at a given offset
Lets look to this syscal closely.

pwrite's definition is as follows;

ssize_t pwrite(int fd, const void *buf, size_t count, off_t offset);

So, now we understand the pwrite. 
Next, we analyze DBWR's pwrite syscalls.

We have similar lines in our strace output.. Like the following;
pwrite(116, "\2\242\0\0\260O\5\0302!\"\204\211\5\33\4/\266\0\0;\0\31"..., 8192, 2851471360) = 8192

When we decode above line according to the syscall's definition , we see ->

DBWR issues write operations to a file, using one of its file descriptor 116 .
It requests to write 8192 bytes (3th value in the arguments = 8192) and its successfully writes all of them. (=8192) --> Another derived data from here is that , the database block size is probably 8k :) 

But what is this file? 
Lets go to /proc filesystem and find out which file it is...
procfs (or the proc filesystem) is a special filesystem in UNIX-like operating systems that presents information about processes and other system information in a hierarchical file-like structure, providing a more convenient and standardized method for dynamically accessing process data held in the kernel than traditional tracing methods or direct access to kernel memory

So we go to /proc filesystem and look to the file descriptors of the DBWR process.

cd /proc/16748/fd
 ls -lrt |grep 116
lrwx------  1 root root 64 Jan  9 15:05 116 -> /erman/erman_33.dbf

So file descriptor with number 116 is pointing to /erman/erman_33.dbf

So in conclusion;

Just looking from the OS layer, we can say that; dbwr is running and writing aggresively to the datafile named /erman/erman_33.dbf..
We did this analysis on the OS layer, as you see. We dont need know anything about the Oracle to make this kind of analysis.

Tuesday, January 7, 2014

Exadata -- Changing DNS Server

Changing the DNS of Exadata means actually changing the DNS of Db nodes, Cell nodes, IB switches and Ethernet switches.
We follow step to accomplish these;
  • Logon to infiniband switches one by one and make the changes in /etc/resolv.conf.. Open /etc/resolv.cof in vi editor, and change the desired dns server or append a line for your new dns server..
Note that : The resolv.conf file is the resolver configuration file. It is use to configure client side access to the Internet Domain Name System (DNS).
  • Logon to Db nodes one by one and make the changes in /etc/resolv.conf..pen /etc/resolv.cof in vi editor, and change the desired dns server or append a line for your new dns server..
  • Logon to one of the Cell nodes, and execute ipconf to change the DNS of Cell Server and ILOM.. In the example above, we add a new dns server( to Exadata.. Note that , Exadata in the Example, has already a dns server (
[root@dm01cel01 oracle.cellos]# /opt/oracle.cellos/ipconf
Logging started to /var/log/cellos/ipconf.log
Interface ib0 is Linked. hca: mlx4_0
Interface ib1 is Linked. hca: mlx4_0
Interface eth0 is Linked. driver/mac: igb/00:21:28:8e:b2:98
Interface eth1 is ... Unlinked. driver/mac: igb/00:21:28:8e:b2:99
Interface eth2 is ... Unlinked. driver/mac: igb/00:21:28:8e:b2:9a
Interface eth3 is ... Unlinked. driver/mac: igb/00:21:28:8e:b2:9b

Network interfaces
Name State IP address Netmask Gateway Net type Hostname
ib0 Linked
ib1 Linked
eth0 Linked
eth1 Unlinked
eth2 Unlinked
eth3 Unlinked
Warning. Some network interface(s) are disconnected. Check cables and swicthes and retry
Do you want to retry (y/n) [y]: n
The current nameserver(s):
Do you want to change it (y/n) [n]: y
Add more nameservers (y/n) [n]: y
Add more nameservers (y/n) [n]: n
The current timezone: Europe/Istanbul
Do you want to change it (y/n) [n]: n
The current NTP server(s):
Do you want to change it (y/n) [n]: n
Network interfaces
Name State IP address Netmask Gateway Net type Hostname
eth0 Linked Management dm01cel01.erman.local
eth1 Unlinked
eth2 Unlinked
eth3 Unlinked
bondib0 ib0,ib1 Private dm01cel01-priv.erman.local
Select interface name to configure or press Enter to continue:
Select canonical hostname from the list below
1: dm01cel01.erman.local
2: dm01cel01-erman.local
Canonical fully qualified domain name [1]:
Select default gateway interface from the list below
1: eth0
Default gateway interface [1]:
Canonical hostname: dm01cel01.erman.local
Timezone: Europe/Istanbul
NTP servers:
Default gateway device: eth0
Network interfaces
Name State IP address Netmask Gateway Net type Hostname
eth0 Linked Management dm01cel01.erman.local
eth1 Unlinked
eth2 Unlinked
eth3 Unlinked
bondib0 ib0,ib1 Private dm01cel01-priv.erman.local
Is this correct (y/n) [y]:
Do you want to configure basic ILOM settings (y/n) [y]:
Loading basic configuration settings from ILOM ...
ILOM Fully qualified hostname [dm01cel01-ilom.erman.local]:
ILOM IP discovery (static/dhcp) [static]:
ILOM IP address []:
ILOM Netmask []:
ILOM Gateway or none []:
ILOM Nameserver or none []:
ILOM Use NTP Servers (enabled/disabled) [enabled]:
ILOM Use NTP Servers (enabled/disabled) [enabled]:
ILOM First NTP server. Fully qualified hostname or ip address or none []:
ILOM Second NTP server. Fully qualified hostname or ip address or none [none]:
Basic ILOM configuration settings:
Hostname : dm01cel01-ilom.erman.local
IP Discovery : static
IP Address :
Netmask :
Gateway :
DNS servers :
Use NTP servers : enabled
First NTP server :
Second NTP server : none
Timezone (read-only) : Europe/Istanbul
Is this correct (y/n) [y]:
Info. Run /opt/oracle.cellos/validations/init.d/saveconfig
Info. Custom changes have been detected in /etc/resolv.conf
Info. Original file will be saved in /etc/resolv.conf.backupbyExadata
Warning. You modified DNS name server.
Ensure you also update the Infiniband Switch DNS server
if the same DNS server was also used by the Infiniband switch.

  • After this operation, restart cell services with the following command;

cellcli -e alter cell restart services all

  • Ensure the cell processes are running and their status are ONLINE ...

cellcli -e list cell detail

  • Repeate this step one by one for each exadata cell...

Monday, January 6, 2014

EBS -- Error: FNDCPASS was not able to decrypt password for user

Changing password, especially Apps user's password is one of the security routines in Apps Dba's Life.
We periodically change the Apps user passwords in our Production Environments, as well as in UAT, TEST and DEV environments according to the securit policies.. Also, we usually change the Apps user's password before we deliver a clone environment..

To change the APPS password , we use the FNDCPASS utility, or AFPASSWD utility if Ebs version is >= R12.1.2.

The usage of FNDCPASS is as follows;
FNDCPASS logon 0 Y system/password mode username new_password
where logon is username/password[@connect]
system/password is password of the system account of that database
username is the username where you want to change its password
new_password is the new password in unencrypted format
example FNDCPASS apps/apps 0 Y system/manager SYSTEM APPLSYS WELCOME
FNDCPASS apps/apps 0 Y system/manager ORACLE GL      GL1
FNDCPASS apps/apps 0 Y system/manager USER   VISION  WELCOME

The thing makes me write this post is, an issue that may appear while changing the APPS password..
To demonstrate ; we execute the FNDCPASS command to change the apps password as below;
FNDCPASS apps/apps 0 Y system/manager SYSTEM APPLSYS APPS

And here it is , look what we see in the log file;

FNDCPASS was not able to decrypt password for user 'ERMAN1' during applsys
password change.
FNDCPASS was not able to decrypt password for user 'ERMAN2' during applsys
password change.
FNDCPASS was not able to decrypt password for user 'APPLSYSPUB' during applsys
password change.
Concurrent request completed successfully
Current system time is 06-JAN-2014 10:47:53

So, by looking to above messages; we can say that FNDCPASS could not actually change the passwords of the application users :ERMAN1,ERMAN2 and the APPLSYSPUB user..
Maybe we can ignore ERMAN1 and ERMAN2, but we cant ignore APPLSYSPUB, as Oracle applications connect uses APPLSYSPUB in the login phase.

The cause of the issue seems a logical corruption or inconsistency in the fnd_user table.
As a solution, we need to reset the passwords of these problematic users beforehand.. This action will fix the unacceptable data related to our users..

Here is the action plan / solution :

Executed following commands;

FNDCPASS apps/apps 0 Y system/manager USER ERMAN1 ERMAN1
FNDCPASS apps/apps 0 Y system/manager USER ERMAN2 ERMAN2
FNDCPASS apps/apps 0 Y system/manager ORACLE APPLSYSPUB PUB

Now, lets change the APPS password again. Note that this time; the command produces a clean log:)

FNDCPASS apps/apps 0 Y system/manager SYSTEM APPLSYS APPS

Application Object Library: Version : 11.5.0

Copyright (c) 1979, 1999, Oracle Corporation. All rights reserved.


Current system time is 06-JAN-2014 13:54:18



Concurrent request completed successfully
Current system time is 06-JAN-2014 13:54:26