[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
RE: How to set new paged result limit [WAS] RE: Paged results flakey - and hdb vs bdb
> Ok, so I think I finally got the format figured out for the new limit
> stuff.
>
> Before my database configuration lines, I have this line:
>
> Sizelimit 30
>
> After my database configuration lines, I have this line:
>
> limits anonymous size.pr=none size.prtotal=none
This is a workaround. Checkout the new fix.
>
> Unless I am still messing things up, here is what I have found:
>
> size.pr=int - works
> size.pr= none - works
> size.pr=noEstimate - works
> size.pr=disabled - doesn't work
This disables the control, i.e. no results are returned, and an error is
issued. What is the behavior you get?
>
> size.prtotal=int - works
> size.prtotal=none - doesn't work
This has been fixed few minutes ago.
>
>
> I think I'm probably still doing something wrong, however, as I don't
> think I am fully understanding how database hard limits, soft limits,
> and the server limits set by the sizelimit keyword interact.
regular searches:
soft limits: activated if no explicit limit is requested (e.g. -z, -l)
hard limits: activated if explicit limit is request (e.g. -z, -l)
unchecked: based on estimated candidate number (regardless of -z, -l)
hard must be >= soft
paged results: by default uses hard limits; regular hard limits
can be overridden by prtotal.
Same applies to time limits, but apart from hard and soft, other
limits are meaningless.
>
> Oh, and there seems to be some mistakes in the man page for slapd.conf -
> In the limits section, whenever it talks about what it will do if the
> requested limit exceeds the hard limit, is says:
>
> If no time limit is explicitly
> requested by the client, the soft limit is used; if the
> requested time limit exceeds the hard limit, an is returned.
>
> I assume that should say "an error is returned"? This happens in
> multiple places throughout the limits discussion.
I'll check that. Thanks.
>
> ps - I'm currently running the code I checked out from HEAD this
> morning.
I fixed the prtotal bug few minutes ago. Sorry for any inconvenience,
and thanks for testing.
p.
--
Pierangelo Masarati
mailto:pierangelo.masarati@sys-net.it