Hello,

I'm writing a application that sets up a Environment with a primary and
several secondary databases. Most of the secondary databasse are set up
with DB_DUPSORT, but I want to avoid duplicates in one of them. The
environment is protected by transactions.

However, when I insert data in the primary database, DB->put returns EINVAL.

For now, all is well since I only have one secondary database which
enforces uniqueness.

But what was I to do if I wanted multiple secondary databases with
uniqueness enforced? How would I find out what secondary database(s)
caused DB->put to return EINVAL?

Are there other and better ways to enforce uniqueness on secondary databases?

--
Frode Nordahl

Hi Frode,

As a first step you should turn on verbose errors as described
elsewhere in this newsgroup and here:

http://dev.sleepycat.com/resources/f...io n%3Dsearch

That should give you a better error message and hopefully help you
identify where the problem is. Let me know if after turning on verbose
errors, you still can't find the source of the problem.

Ron

On 2005-11-28 17:10:24 +0100, ron.cohen (AT) gmail (DOT) com said:
That
Thank you for your reply!

I think I was a bit too much inside my mind when I phrased the first
question :-)

My test program inserts two records with different primary key, and the
same data. I have multiple secondary databases, one of them is not
configured to support duplicates. So this will obviously fail.

When turning on error logging I get the following:
"Put results in a non-unique secondary key in an index not configured
to support
duplicates"

This is what I expected to get.

As of now I only have one secondary database not configured to support
duplicates, so I check for EINVAL on DB->put, and handle things
accordingly. But I may run into a situation where DB->put returns
EINVAL for a different error, or I may need multiple secondary
databases that enforce uniqueness.

My question is: Is it possible to identify which secondary database the
key allready exists in?

--
Frode Nordahl