Re: Introducing: java-rrd
"Peter Stamfest" <[email protected]>
| Newsgroups | gmane.comp.db.rrdtool.devel |
|---|---|
| Message-ID | <OF548244BE.7EB93337-ONC12577A6.002CDED0-C12577A6.002E4FD7@domino.stamfest.net> |
Tobias Oetiker <[email protected]> schrieb am 22.09.2010 06:40:53: > Yesterday Peter Stamfest wrote: > > > Hello! > > > > This is the first public release of java-rrd [1]. It is (yet another) > > attempt on a java RRD interface. However, it uses the pipe interface to > > the rrdtool proper. It is therefore similar to how the RRDp perl package > > works. It has the additional advantage of being 100% compatible with the > > rrdtool on whatever platform it is run on, because it uses the original > > rrdtool. [...] > > Development of java-rrd is done using mercurial. You can clone a repo from > > > > > > http://oss.stamfest.net/java-rrd-hg/ > > > > A tarball of the current state can be found at > > > > http://oss.stamfest.net/java-rrd/java-rrd-18.tar.gz > > > > Any input/feedback is welcome. Patches are welcome. Documentation patches > > are even more welcome. > > now if I only was using more java :-) great stuff ... maybe the > opennms people could use this ! They could certainly make use of it. Currently I have only implemented the "create", "update", "tune" and "graphv" commands, but the basic infrastructure for all other commands is there. I have also implemented a simple pooling service, so this is actually multi-threading capable for EVERY implemented rrdtool command [and not just for those with explicit support in librrd_th] and a maximum-request count per client to avoid any negative effects of a long running rrdtool (eg. leaks of any kind). It also uses an internal form of "transaction" which allows for the use of the "mkdir", "cd" and "ls" commands. I have actually implemented this for a very long running personal project of mine - a documentation system for networks that can also do a lot of (distributed!) monitoring (including Nagios configuration support, NRPE deployment and RRD graphing). This was partly because I really didn't like OpenNMS (sorry guys!). But I think that a common Java based RRD interface is something our little segment of the world misses quite badly. I also want to say, that the pipe interface seems to be THE (most) correct interface for rrdtool integration in every language. Many, many thanks for this interface. If only I would have realized this earlier! OTOH, there might be some improvements for this interface as well... But before I submit anything, first let's see how this iteration of Java RRD support turns out. peter _______________________________________________ rrd-developers mailing list [email protected] https://lists.oetiker.ch/cgi-bin/listinfo/rrd-developers