Hi, Lennart,<div><br></div><div>Thanks for your informaiton.</div><div><br></div><div>Looked into it deeper, it might not be a bug of Avahi, just hundred hosts in the same network which periodically sending Multicast package might be a heavy load to avahi-daemon, which is running in an embedded system.</div> <div><br></div><div>In my case, seems like too many network traffic causes too many avahi internal time events been created, between a very tiny interval. Just wonder whether it's possible to change below place to optimize this condition in my case (given that I don't need to consider about Apple Bonjour compliance test right now)?</div> <div><br></div><div>#define AVAHI_RESPONSE_JITTER_MSEC 500</div><div><br></div><div>Thanks!</div><div><br></div><div>Leyn<br><br><div class="gmail_quote">2009/6/29 Lennart Poettering <span dir="ltr"><<a href="mailto:lennart@poettering.net">lennart@poettering.net</a>></span><br> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im">On Wed, 24.06.09 12:11, chen leyn (<a href="mailto:leyncl@gmail.com">leyncl@gmail.com</a>) wrote:<br> <br> > Hi, there,<br> > I'm recently working on porting Avahi (version 0.6.22) to an embedded Linux<br> > (kernel 2.6.17) platform. The job is almost done and we already achieved all<br> > the functionalities we want. However, during further testing, we found in<br> > some lab that there are heavy Bonjour network traffic (many Mac OS hosts),<br> > Avahi-daemon will eat CPU time periodically.<br> ><br> > We use strace to find what is happening, and here is the data, seems like<br> > the timeout value which avahi calculated for next poll operation is always 0<br> > for a while, not sure whether that's expect behavior and whether such tight<br> > poll loop eat CPU time. Also want to mention that, after we plugged out the<br> > network cable, the symptom disappear, after plugged in back, appear again.<br> > Even after kill avahi-daemon and restart it, problem is still there.<br> ><br> > Any help we can get from here? Thanks in advance!<br> <br> </div>No clue, except maybe that gdb or even oprofile might be a more useful<br> tool here.<br> <br> I fear you have to track that one down yourself, as I haven't had any<br> report about this before.<br> <br> Sorry,<br> <br> LennartLennart<br> <font color="#888888"><br> --<br> Lennart Poettering Red Hat, Inc.<br> lennart [at] poettering [dot] net<br> <a href="http://0pointer.net/lennart/" target="_blank">http://0pointer.net/lennart/</a> GnuPG 0x1A015CC4<br> _______________________________________________<br> avahi mailing list<br> <a href="mailto:avahi@lists.freedesktop.org">avahi@lists.freedesktop.org</a><br> <a href="http://lists.freedesktop.org/mailman/listinfo/avahi" target="_blank">http://lists.freedesktop.org/mailman/listinfo/avahi</a><br> </font></blockquote></div><br></div>