ando@OpenLDAP.org wrote:
add monitor support for unindexed attributes (need to manually #define BDB_MONITOR_IDX)
This patch adds an operational attribute, olmBDBNotIndexed, that contains information about search operations directed to this database and using unindexed attributes in the filter. As a result, the monitor entry for the database will contain attributes like
olmBDBNotIndexed: <attr>#<count>#<index>
where <attr> is the attribute, <count> is the number of occurrences of that attribute in a search filter that could not use indexes, and <index> is the type of filters that were used, with the syntax used for the "index" directive. For example,
olmBDBNotIndexed: mobile#91#eq olmBDBNotIndexed: telephoneNumber#1876#eq,sub
To enable, one needs to explicitly #define BDB_MONITOR_IDX.
p.
Ing. Pierangelo Masarati OpenLDAP Core Team
SysNet s.r.l. via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it --------------------------------------- Office: +39 02 23998309 Mobile: +39 333 4963172 Email: pierangelo.masarati@sys-net.it ---------------------------------------
<quote who="Pierangelo Masarati">
ando@OpenLDAP.org wrote:
add monitor support for unindexed attributes (need to manually #define BDB_MONITOR_IDX)
This patch adds an operational attribute, olmBDBNotIndexed, that contains information about search operations directed to this database and using unindexed attributes in the filter. As a result, the monitor entry for the database will contain attributes like
olmBDBNotIndexed: <attr>#<count>#<index>
where <attr> is the attribute, <count> is the number of occurrences of that attribute in a search filter that could not use indexes, and <index> is the type of filters that were used, with the syntax used for the "index" directive. For example,
olmBDBNotIndexed: mobile#91#eq olmBDBNotIndexed: telephoneNumber#1876#eq,sub
To enable, one needs to explicitly #define BDB_MONITOR_IDX.
Is it possible to automatically creat indexes via slapd at runtime?
For example, some of the above hit a certain trigger level that is configurable, then an index gets created?
Just wondering...
Gavin.
p.
Ing. Pierangelo Masarati OpenLDAP Core Team
SysNet s.r.l. via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it
Office: +39 02 23998309 Mobile: +39 333 4963172 Email: pierangelo.masarati@sys-net.it
--On Tuesday, August 07, 2007 10:11 PM +0100 Gavin Henry ghenry@suretecsystems.com wrote:
To enable, one needs to explicitly #define BDB_MONITOR_IDX.
Is it possible to automatically creat indexes via slapd at runtime?
For example, some of the above hit a certain trigger level that is configurable, then an index gets created?
You can do this using back-config, and some people don't want particular things indexed, so automatically doing it is a bad idea.
--Quanah
-- Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration
<quote who="Quanah Gibson-Mount">
--On Tuesday, August 07, 2007 10:11 PM +0100 Gavin Henry ghenry@suretecsystems.com wrote:
To enable, one needs to explicitly #define BDB_MONITOR_IDX.
Is it possible to automatically creat indexes via slapd at runtime?
For example, some of the above hit a certain trigger level that is configurable, then an index gets created?
You can do this using back-config, and some people don't want particular things indexed, so automatically doing it is a bad idea.
Of course, I forgot. I did it the other day ;-)
--Quanah
-- Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc
Zimbra :: the leader in open source messaging and collaboration
Gavin Henry wrote:
Is it possible to automatically creat indexes via slapd at runtime?
Yes.
For example, some of the above hit a certain trigger level that is configurable, then an index gets created?
Yes, although perhaps not always advisable.
Just wondering...
I also see this commit as a step towards a self-tuning database (like the initial idea of publishing Berkeley DB stats via back-monitor), but I don't think it's just around the corner.
p.
Ing. Pierangelo Masarati OpenLDAP Core Team
SysNet s.r.l. via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it --------------------------------------- Office: +39 02 23998309 Mobile: +39 333 4963172 Email: pierangelo.masarati@sys-net.it ---------------------------------------
<quote who="Pierangelo Masarati">
Gavin Henry wrote:
Is it possible to automatically creat indexes via slapd at runtime?
Yes.
Yes I forgot, via cn=config.
For example, some of the above hit a certain trigger level that is configurable, then an index gets created?
Yes, although perhaps not always advisable.
Uderstood.
Just wondering...
I also see this commit as a step towards a self-tuning database (like the initial idea of publishing Berkeley DB stats via back-monitor), but I don't think it's just around the corner.
Thanks. Did we have anymore thoughts on a 2.4 beta? I know Howard is quite busy just now, but I seem to remember a few things waiting to be committed from ralf?
Gavin.
p.
Ing. Pierangelo Masarati OpenLDAP Core Team
SysNet s.r.l. via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it
Office: +39 02 23998309 Mobile: +39 333 4963172 Email: pierangelo.masarati@sys-net.it
Gavin Henry wrote:
Thanks. Did we have anymore thoughts on a 2.4 beta? I know Howard is quite busy just now, but I seem to remember a few things waiting to be committed from ralf?
I added (rudimentary) back-config support to slapd-relay(5) and slapo-rwm(5), the big pieces that were missing. I'm still in trouble with slapd-meta(5), it's not going to be as easy. I don't have any further piece 'round for 2.4.
p.
Ing. Pierangelo Masarati OpenLDAP Core Team
SysNet s.r.l. via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it --------------------------------------- Office: +39 02 23998309 Mobile: +39 333 4963172 Email: pierangelo.masarati@sys-net.it ---------------------------------------
Pierangelo Masarati wrote:
Gavin Henry wrote:
Thanks. Did we have anymore thoughts on a 2.4 beta? I know Howard is quite busy just now, but I seem to remember a few things waiting to be committed from ralf?
I added (rudimentary) back-config support to slapd-relay(5) and slapo-rwm(5), the big pieces that were missing. I'm still in trouble with slapd-meta(5), it's not going to be as easy. I don't have any further piece 'round for 2.4.
Ok. I'm just trying to envision a deadline so I can do a big push no the empty doc sections in HEAD.
Gavin.
-
Gavin Henry -
Pierangelo Masarati -
Quanah Gibson-Mount