More on ORA-08177: can't serialize access for this transaction

Lionel Samuel lionel.samuel01 at gmail.com
Wed Sep 13 21:25:30 EDT 2017


>I would, if I were you, generate my persistent IDs and not store them. If
the first generation of values in the database >are being produced by the
computed strategy than switching to it would be a reasonable and advisable
thing to do >with no external impact.

If I understood correctly -- with current setup below -- you are
recommending to uncomment 1st line below --- and comment out 2nd line?

#idp.persistentId.computed = shibboleth.ComputedPersistentIdGenerator
idp.persistentId.generator = shibboleth.StoredPersistentIdGenerator
idp.persistentId.dataSource = OracleSource
idp.persistentId.sourceAttribute = UID
idp.persistentId.salt = 000

On Wed, Sep 13, 2017 at 6:03 PM, Cantor, Scott <cantor.2 at osu.edu> wrote:

> On 9/13/17, 8:53 PM, "users on behalf of Lionel Samuel" <
> users-bounces at shibboleth.net on behalf of lionel.samuel01 at gmail.com>
> wrote:
>
> > But still getting errors - now obviously with 'Error retryable, but
> retry limit exceeded' -- any tips on options?
>
> Don't use a databse, or use a database that supports transactions
> correctly. You might also have timeouts set too low and a database that's
> not adequate to get requests in and out fast enough to avoid conflicts.
> Even if the locking wasn't broken, it might still timeout waiting for the
> table.
>
> I would, if I were you, generate my persistent IDs and not store them. If
> the first generation of values in the database are being produced by the
> computed strategy than switching to it would be a reasonable and advisable
> thing to do with no external impact.
>
> I would imagine the only chance Oracle has to work is if you plugin all
> your own SQL, which is a substantial amount of work but there are
> properties you can set for every SQL statement it issues. Oracle has a
> proprietary syntax for doing selects that take table locks, which is what
> it's supposed to do already. I have no idea if it would help, nobody's
> tried it that I know of.
>
> -- Scott
>
>
> --
> To unsubscribe from this list send an email to
> users-unsubscribe at shibboleth.net
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://shibboleth.net/pipermail/users/attachments/20170913/88a4f216/attachment.html>


More information about the users mailing list