Home / user-8321

Tim Büthe's questions

Martin Hope
Tim Büthe
Asked: 2009-10-31 04:35:59 +0800 CST
  • 1

I just upgraded to Ubuntu 9.10 and now my db2 is not starting anymore. The error message reads as follows:

db2start: error while loading shared libraries: libstdc++.so.5: cannot open shared object file: No such file or directory

I guess, this is, because Ubuntu now uses version 6 as I can see in Synaptic. Is there a way, to tell DB2 to use the newer Version or do I have to Install the 5er Version? Will this somehow conflict with the 6er?

ubuntu
Martin Hope
Tim Büthe
Asked: 2009-10-12 23:48:37 +0800 CST
  • 3

I have the following openldap server configuration:

access to attrs=userPassword
    by self write
    by anonymous auth
    by set="[cn=users,ou=Group,dc=my-company,dc=de]/memberUid & user/uid" write
    by * none

# Allow everybody adding and changing Contacts
access to dn.subtree="ou=Contacts,dc=my-company,dc=de"
    by set="[cn=users,ou=Group,dc=my-company,dc=de]/memberUid & user/uid" write 
    by * read

access to *
    by self write
    by dn.base="cn=admin,dc=my-company,dc=de" write
    by set="[cn=sysadm,ou=Group,dc=my-company,dc=de]/memberUid & user/uid" write
    by * read

What it should do is this:

  1. allow everybody to change his own password
  2. allow all users to add and change contacts
  3. allow all in group "sysadm" to change everything

The problem is, the sysadms can not change any user password. Any hints?

ldap
Martin Hope
Tim Büthe
Asked: 2009-08-29 01:18:35 +0800 CST
  • 0

I know, there is the database configuration parameter "Database collating sequence" for setting the collation in db2. But since this effects the whole database and can only be set when creating it, I'm looking for the possibility for setting this for a single table or column. Alternatively, setting the collation in the SQL statement, like you could do on Microsoft SQL Server:

SELECT * FROM table ORDER BY col1 COLLATE Latin1_General_CI_AS

Is this possible in DB2? How?

Secondly, how will changing the collation for either for the whole db, a table, or within a query effect the performance?

db2
Martin Hope
Tim Büthe
Asked: 2009-06-06 02:11:15 +0800 CST
  • 6

I face some trouble with IBM DB2's auto-increment columns. At first, all my columns were defined as GENERATED ALWAYS, but since I had trouble with this when using the "db2 import ..." command, I changed them to GENERATED BY DEFAULT. This is necessary, sinceI need the IDs to be consistent, because other tables reference them. So using "db2 import ... modified by identityignore ..." isn't an option.

When I now import data, the IDs are inserted correctly, but everytime I do this, I have to remember to set a new start for the auto-increment column by getting the highest Id+1 and alter the column like this:

SELECT MAX(mycolumn)+ 1 FROM mytable;
ALTER TABLE mytable ALTER COLUMN mycolumn RESTART WITH <above_result>;

If I forget this, an Insert-Statement will fail with an duplicate PK error, since the auto-increment column is the primary key.

So my question is: Is there a way to find the next value for an auto-increment column, so I could write Statements that would check, if this value is less then the SELECT MAX and needs to be set?

Or: Isn't this whole thing as complicated as it seems to me? Could I somehow import data, preserving the IDs and have the auto-increment column still working as expected?

sql database db2