[AGENT++] Problem with READ-CREATE tables...

[Fedja Jeleskovic]

> I recommend to move any code that returns an error code
> from the commit_set_request to the prepare_set_request
> method (in your example this seems to be true for most
> of the code).

Thanks, I will move all of that in the prepare_set_request method. I
am assuming that includes whole switch block with the rowDestroy code
as well.

> Reindexing the rows in a SNMP table is not a good idea
> and I strongly recommend to not implement it, because
> managers rely on indexing and if the same row gets a new
> index between two poll cycles, the manager may not be
> able to find or identify the row it is supposed to poll.

Hmm, well, the problem here is that my storage gets reindexed by it
self as soon as any change is done in there. So, what happens is that
after removal of the row (let say 3rd row of 4 in the table), row with
index 4 is not accessible since there is nothing in the storage with
that index.

Any idea how to go around that?


Thanks


Fedja