Hi Moshe
this is not possible as the problem is that tcpdump killed with -9 leaves the 
queue in an inconsistent state (same happens if it crashes)

Alfredo

> On 18 Jul 2016, at 15:30, Moshe Danielli <[email protected]> wrote:
> 
> Got it - Thanks !
> 
>  
> A followup question - is there an option to manually release the queue, in 
> case kill -9 was indeed used ?
>  
> Thanks again,
>  
> Moshe
>  
> 
> On Jul 18, 2016 14:44, Alfredo Cardigliano wrote:
> Hi Moshe
> killing tcpdump with -9 prevents it from releasing the queue correctly, thus 
> you are no more able to open it. This is an expected behaviour in zero-copy 
> mode.
> 
> Alfredo
> 
> > On 18 Jul 2016, at 13:13, Moshe Danielli wrote:
> > 
> > 
> > Hello Guys,
> > 
> > I've been experiencing a problem where I execute the following sequence of 
> > commands:
> > 
> > 1. nohup tcpdump -i zc:99@0 &
> > 2. kill -9 tcpdump_pid
> > 3. tcpdump -i zc:99@0 -> this command yields the following error: "tcpdump: 
> > zc:99@0: No such device exists"
> > 
> > It's important to note that, killing tcpdump, without using the KILL signal 
> > (hitting CTRL-C for example or just using: "kill tcpdump_pid") will let me 
> > reuse that cluster node.
> > 
> > I'm curious as to why that's the current behavior.
> > 
> > Thanks in advance
> > 
> > 
> > _______________________________________________
> > Ntop-misc mailing list
> > [email protected]
> > http://listgateway.unipi.it/mailman/listinfo/ntop-misc
> _______________________________________________
> Ntop-misc mailing list
> [email protected]
> http://listgateway.unipi.it/mailman/listinfo/ntop-misc
> 
> 
> 
> 
> _______________________________________________
> Ntop-misc mailing list
> [email protected]
> http://listgateway.unipi.it/mailman/listinfo/ntop-misc
_______________________________________________
Ntop-misc mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-misc

Reply via email to