[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[LDM #FXW-644009]: question about allow patterns



Hi Karen,

> Could you clarify for me the changes that have been made to the allow
> pattern that let you narrow down what can be transmitted to downstream
> machines?
> 
> The unidata ldmd.conf webpage says I can do:
> 
> ALLOW feedtype hostIdEre [Ok_pattern [Not_pattern] ]
> 
> so I'm thinking I can do something like:
> 
> ALLOW  ANY 129.15.109.3 [[KCRI]]
> 
> if I want the 129.15.109.3 downstream machine to get everything except
> data-product identifiers which match KCRI.

To do what you want, the ALLOW-line should be

    ALLOW  ANY  129.15.109.3  .*  ^KCRI

assuming that the "KCRI" string appears at the start of the
product-identifier.  If you want to prevent products that
have "KCRI" anywhere in their product-identifier from being
sent, then remove the caret (^) from the not-pattern.

If you use the not-pattern, then the OK-pattern must exist
and would, typically, be ".*" (without the quotation marks).

The square brackets are merely the conventional UNIX way of
indicating an optional component: they shouldn't appear in
the ALLOW-line (where they would mean something quite
different to the regular expression library).

> Basically, we are going to be doing some tests on a new data stream, and
> the feedtype is the same as most of our other streams.  However, we
> don't want all of the downstream users to get the new test data until
> it's been *blessed*.
> 
> I had originally contemplated using a different feedtype during the
> testing, but that's a lot harder (and involves a lot more red tape) than
> modifying the ldmd.conf on the intermediary machine.
> 
> If you can think of some better way to do this, or if I'm missing
> something in my reasoning, please let me know.

Contact me if you have any questions.

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: FXW-644009
Department: Support LDM
Priority: Normal
Status: Closed