cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Did you get called away in the middle of writing a post? Don't worry you can find your unfinished post later in the Drafts section of your profile page. X

R9.1 M040 - R10.0 M030 issues

ssenthil
1-Visitor

R9.1 M040 - R10.0 M030 issues

We consistently hit the following issues in our rehearsal migrations.



R9.1 M040 - R10.0 M030 (with and without R10 M030 PS1 generic patch)



1. Upgrade manager removes the ldap pending users node in "Configure
Database and LDAP" step. But when executing standalone migrators, it is
failing due to the non existing pending users node. Once importing the node,
upgrade manager moves on.



2. Upgrade manager is failing when attempting to set default values for the
columns TASKURI, RESPONSEMETAINFOPATH, ESITARGETNUMBER in ESITARGET table.
When those columns are manually added, upgrade manager is progressing to
next steps. But OOTB ESITARGET table doesn't have those three columns.



3. When upgrade manager is attempting to start method server, method server
is failing to start with the following exception: wt.util.WTException:
FILENAMEUSN_SEQ's current value is lesser than UniqueSequenceNumber. When
the interval of sequence "FILNAMEUSN_SEQ" is set to 1000(default is 100),
the method server is coming up. Advance config guide recommends two alter
sql statements but they don't have any influence on this behavior.



TS is taking a look at those issues. Anyone else noticed the above issues.



Thanks in advance.





Thank you and have a great time.

Best Regards

Swamy Senthil

Principal Solutions Architect, Swasen Inc

<">mailto:swamy.senthil@swasen.com(Email)>
swamy.senthil@swasen.com(Email); 909 800 8423(M); 973 324 2729(W); 866 908
6561(F);



4 REPLIES 4


Swamy,On the third issue, the migrator is incorrect and is to be corrected in 10.1 UpgradeManager. FILENAMEUSN_SEQ is created during the 10.0 Upgrade and initial setting of LAST_NUMBER has been out of sync with the LAST_NUMBER of UniqueSequenceNumber (ID_SEQUENCE) on all our 9.1-10.0 migrations. When the MethodServer fails to start, in SQLselect * from all_sequences where sequence_name in ('FILENAMEUSN_SEQ','ID_SEQUENCE')
and locate the LAST_NUMBER of the ID_SEQUENCEtake that number and place in on the 'START WITH XXXXXXXXXX' below.



drop SEQUENCE
FILENAMEUSN_SEQ


CREATE SEQUENCE FILENAMEUSN_SEQ


INCREMENT BY 100


START WITH 598558101


MAXVALUE 999999999999999999999999999


MINVALUE 1




restarted UGmanager.
-

Hello RJ

thanks for confirming the issue and providing a fix. Did you face other two
issues (1 and 2) that we are facing.



Thanks in advance



Swamy Senthil

Swasen Inc.



Swamy, Never encountered the problem with pending LDAP users (all tasks in the Config Database and LDAP completed 100%).ESI was not implemented in 9.1 (will be in 10.1) so it was not a problem.



-

I have not had these issues on 10.0 upgrades, but have experienced similar issues in 9.0/9.1 upgrades.


For you ldap pending users missing issue - where are those supposed to come from? Is the Upgrade Manager responsible to move them, another utility or a manual process? Those have been moved manually in upgrades I've done in the past. Be sure you know precisely the step they are missed on and look in the logs to see if there is a reason for failure if it is indeed a programatic failure.


For the missing columns, and default values, I have had similar issues on previous upgrades, the columns should be added during a upgrade step like 'upgrade schema'. You should see failures in the logs, they will be helpful to uncover why the columns were not addeds.


In my experience it is not dangerous to proceed with manual fixes in upgrade rehearsals, but I prefer patches before production to be used on my penultimate upgrade rehrearsal to make sure they are good (if in fact patches are the fix).


/pete

Announcements


Top Tags