Wednesday, December 6, 2017

WebSphere Application server: how reset a password from file registry using wsadmin

Sometimes it happens.. I had to login to a WebSpere DMGR but the password of the local administrator wasn't working as expected and there wasn't LDAP configured.

One of the easy way you have to get back your Cell is reset the password of the local admin using wsadmin

  • open a command promp on  dmgr profile bin
  • wsadmin -conntype NONE
  • $AdminTask changeFileRegistryAccountPassword {-userId waslocal -password ab12c34de}
  • $AdminTask change

Restart the JVM and you will be able to login with your waslocal userId and ab12c34de password. After this task perform a syncNode on every Node to synchronize this change.

If you need more details please read this wiki page.

Wednesday, November 29, 2017

IBM Connections Engagement Center announced for IBM Conncetions Cloud: dont' miss the related Webinar next friday !

IBM Connections Engagement Center ( ICEC ) is a plattform that brings to IBM Connections some wcm feature we could usually find working with content management platfroms:
  • Create and customize pages
  • Add content and applications with widgets and APIs
  • Personalize content based on employee directory data
  • Simplify publishing
  • Integrate and administrate quickly and efficiently
ICEC is already available for on-prem enviroment but now IBM has announced the availability for IBM Connections cloud too , click here for  the related announcement letter.



The release date should be very soon, probably during 30th November, here you could also see the related page on IBM Connections cloud wiki what's new.


ICEC will not be included in current IBM Connections cloud license and require a little extra fee.
IBM has organized a Webinar to provide all the related updates and news for

Fri 1 December 2017 14:30 – 15:00 CET

follow the link below to the registration page and reserve your seat !
http://ibm.biz/EngagementCenterDec2017

Sunday, November 26, 2017

WebSphere Portal how to: Recover a WPSDB on derby with SQL status XJ040

Some days ago I've received a new ticket from one of my customers with a WebSphere portal unable to start due some issue on WPSDB.

This was a dev environment so WPSDB was resident on local derby instance

inside  SystemOut.log this was the relevant error during Portal start

Failed to start database 'C:/IBM/WebSphere/wp_profile/PortalServer/derby/wpsdb' with class loader [email protected], see the next exception for details.DSRA0010E: Stato SQL = XJ040, Codice errore = 40.000
Failed to start database 'C:/IBM/WebSphere/wp_profile/PortalServer/derby/wpsdb' with class loader [email protected], see the next exception for details.
Recovery failed unexpected problem log record is Not first but transaction is not in transaction table : 40980828.
ERROR XSLAO: Recovery failed unexpected problem log record is Not first but transaction is not in transaction table : 40980828.


Inside websphere setup there are some derby libraries available under
WASHOME\derby\lib

so to verify the WPSDB status we have copied the WPSDB folder in a temporary directory than we have started the troubleshooting:


java -cp "derby.jar;derbytools.jar" org.apache.derby.tools.ij


this prompt was the response
ij>

here we tried to connect to our copy
ij> connect 'jdbc:derby:C:/temp/wpsdb';

the connections worked so we checked the table status with the following command
SELECT schemaname || '.' || tablename as TableName,SYSCS_UTIL.SYSCS_CHECK_TABLE(schemaname, tablename) AS OK FROM sys.sysschemas s, sys.systables t WHERE s.schemaid = t.schemaid   and t.tabletype = 'T'   ;

tables status looked good so this point us to derby tranlog corruption.
To have a new log with correct file structure we have created a new empty derby db with following command
ij> connect 'jdbc:derby:MyDbTest;create=true';
after that we have copied the  log/  from new db to log/ of  wpsdb (all steps with releated backup) and after have copied back the new wpsdb folder the Portal was able to start again !

I like to thank  Andrea Grassi for his blogpost about derby corrupted database recovery guide.

Tuesday, November 21, 2017

IBM Connections 6.0 CR1 available for download

IBM has just released on fixcentral the new CR1 for Connections 6.0 that include all fixes released until now (Day 1 , June Update) with a lot of more update and optimizations

Below some useful links and info:




WebSphere Appserver support is still to  8.5.5 FP11 ,  if you needs any others details I encourage you to read the upgrade strategy page.

Thursday, November 16, 2017

WebSphere How To: fix errore ADMG0011E while Application Server deletion

Working on a WebSphere 8.5.0.2 environment I was unable to delete an Application server getting the error ADMG0011E:

An error occurred while deleting the server. ADMG0011E: An unexpected exception occurred com.ibm.websphere.management.exception. AdminException: Exception in removeTargetFromCUs()..


Inside DMGR and NodeAgent logs I wasn't able to find anything related but inside DMGR FFDC, folder I got a log generated during the deletion time frame with following exceptions:

Caused by: com.ibm.wsspi.management.bla.op.OpExecutionException: CWWMH0121E: Could not read configuration data for composition unit "WebSphere:cuname=<Application_Name>,cuedition=BASE". Error data: com.ibm.wsspi.management.bla.op.OpExecutionException: Enterprise application data for composition unit "WebSphere:cuname=<Application_Name>,cuedition=BASE" could not be read. at com.ibm.ws.management.bla.model.CompositionUnitConfig.read(CompositionUnitConfig.java:293)

After some research  I've found this question on  DeveloperWorks and than I've fixed the issue following this steps:

  1. Deployement Manger stop
  2. check inside WAS_INSTALL_ROOT/profiles/Dmgr_ProfileName/config/cells/Cell_Name/applications if an  Application_Name.ear folder with same FFDC name Application_Name was present
  3. If the folder on step 2 was missing check inside WAS_INSTALL_ROOT/profiles/Dmgr_ProfileName/config/cells/Cell_Name/blas and than delete  Application_Name if present
  4. check inside WAS_INSTALL_ROOT/profiles/Dmgr_ProfileName/config/cells/Cell_Name/cus if the  Application_Name folder is present than delete
  5. Deployement Manger start
  6. Full Nodes Sync from WebSphere console
  7. Tried again to delete the  Application server and it worked!

Thanks to  Jeanette Daly for the detailed response provided at this page .

Friday, November 10, 2017

ICON UK Events: Around the Table - Nov 17 : Domino and Notes 10, Verse On Prem 1.0.3, Domino 9.0.1 Fp10 and more ! Replay available

Today there was a really interesting webinar from the monthly  ICON UK around the table webinar about IBM di Domino 10 announcement and others interesting topics.


I really enjoyed the Q&A part with
This are some of the response:

Domino2025jam 
Jam will be mostly about the future after Domino10 but IBM will listen about request to better understand  what community and customers needs and  to prioritize the related developments
.
9.0.1 FP10
FP10 will be released probably during January 2018

Beta program Domino10
Yes ! There will be a beta program ! 

Domino 10 e Notes 10
Yes ! We will also  have Notes 10

Verse On prem 1.0.3 disponibile seconda metà 
Probably during December 2017 (second half) will be released VOP 1.0.3 with mail&calendar delegation and maybe Sametime integration .

Xpages
There is no no intent to discontinue but the future is under reviews .

below the webinar replay from   ICON UK youtube channel   (thanks guys for your work! )




Wednesday, November 8, 2017

IBM WebSphere 9 with Java 8.0 Service Refresh 5 : how to fix sync issue when LTPA token become invalid

During last weeks I was working on  WebSphere Portal 9 environment who runs over a WebSphere Application server 9.0 FP4 with last JVM available 8.0 SR5 .  Every day after some hours I've noted the nodes goes in unsynched status without a clear reason and after a manual SyncNode the syncronization will be restablished until the next day.

After some investigation I noted sync issue was started at the end of the LTPA token with the following errors:

Node Agent:
[11/3/17 17:05:35:849 CET] 0000006b AdminServiceI 3   invoke method throws exception
                                 com.ibm.websphere.management.exception.AdminException: javax.management.JMRuntimeException: ADMN0022E: Access is denied for the isNodeSynchronized operation on NodeSync MBean because of insufficient or empty credentials.

DMGR
[11/3/17 17:05:52:651 CET] 00000105 SecurityHelpe >  isCertPathValidatorException Entry
                                 com.ibm.websphere.security.auth.WSLoginFailedException: Validation of LTPA token failed due to invalid keys or token type.


I've open a PMR and after I've send the log   ( remember to follow the Must gather steps when you set logs to WAS support) IBM support has recognized a similar behavior on  known APAR

IJ00732: JAVA JIT: INCORRECT RESULTS FOR PATTERNS OF SHIFTED AND OR'D BYTES
http://www.ibm.com/support/docview.wss?uid=swg1IJ00732
Due to this JIT failure, sudden LTPA toke validation issue happens and it results into the  synchronization issue.

Would it be possible to see if the issue can be avoided by the JVM entiry?
System administration > Deployment manager > Process definition  > Environment Entries
Name: TR_DisableIORByteSwap
Value: 1

I've followed the suggested workaround and the issue was fixed !
A new JVM 8.0.5.5 now scheduled for Nov 13th  will fix this issue.