On Nov 15, 2005, at 7:11 PM, Herbert Poetzl wrote:
> On Mon, Nov 14, 2005 at 03:09:32AM -0600, Drew Lippolt wrote:
>>
>> QUESTION: what is the current story with multicast support for both
>> sending and receiving multicast traffic?
>>
>> BACKGROUND:
>>
>> trying to get tomcat clustering working in vserver.
>>
>> http://tomcat.apache.org/tomcat-5.5-doc/cluster-howto.html
>>
>> DETAILS:
>>
>> [root@app2 opt]# cat /proc/version
>> Linux version 2.4.30-vs1.2.10 (root@app1.moverotech.com) (gcc version
>> 3.2.3 20030502 (Red Hat Linux 3.2.3-52)) #1 Wed Aug 10 01:27:44
>> CDT 2005
>>
>> [root@app2 opt]# grep MULTICAST /boot/config-2.4.30-vs1.2.10
>> CONFIG_IP_MULTICAST=y
>>
>> tomcat 5.5.12
>>
>> i'm running debian sarge vservers on redhat enterprise linux 3.0
>> boxes at rackspace, using a stock kernel.org kernel with required
>> tweaks for my hardware this is a production environment which has
>> been supporting many apps beautifully for 3 months.
>>
>> the basic idea is that i have multiple app layer boxes, which i want
>> to distribute tomcat clusters. single instance of tomcat, running in
>> a single vserver, per real host, per application. so if i have 3
>> apps, on 3 real servers, i'd have 9 total vservers across 3
>> clusters. i'm not even getting that far. my test setup is 2 real
>> hosts, each with one vserver with an 'out of the box' tomcat cluster
>> config. the tomcat instances aren't finding each other.
>>
>> the stacktraces i'm getting on tomcat --shutdown-- are as follows,
>> which doesn't look all that interesting, its in code looking for
>> incoming connections on a tcp port, which never arrive, since the
>> mcast conversation never happens (ReplicationListener.java:130):
>>
>> SEVERE: Unable to process request in ReplicationListener
>> java.nio.channels.ClosedSelectorException
>> at sun.nio.ch.SelectorImpl.lockAndDoSelect(SelectorImpl.java:
>> 55)
>> at sun.nio.ch.SelectorImpl.select(SelectorImpl.java:70)
>> at org.apache.catalina.cluster.tcp.ReplicationListener.listen
>> (ReplicationListener.java:130)
>> at org.apache.catalina.cluster.tcp.ClusterReceiverBase.run
>> (ClusterReceiverBase.java:394)
>> at java.lang.Thread.run(Thread.java:534)
>>
>>
>> THINGS I"VE TRIED:
>>
>> * add multicast ip to IPROOT. this just causes barf messages at
>> vserver startup
>>
>> Starting the virtual server v208
>> Server v208 is not running
>> SIOCSIFADDR: Invalid argument
>> SIOCSIFFLAGS: Cannot assign requested address
>> SIOCSIFNETMASK: Cannot assign requested address
>> SIOCGIFADDR: Cannot assign requested address
>> SIOCSIFBROADCAST: Cannot assign requested address
>> SIOCSIFBRDADDR: Cannot assign requested address
>> SIOCSIFFLAGS: Cannot assign requested address
>> ipv4root is now 192.168.1.208 228.0.0.4
>
> how did you add it?
tried a few different ways.
IPROOT="192.168.1.237 228.0.0.4"
IPROOT="192.168.1.237 228.0.0.4/224.0.0.0"
>
>> * enabled NET_ADMIN and NET_BROADCAST. this offers no change
>
> well, NET_ADMIN is what you probably need for
> multicasting, NET_BROADCAST should suffice for
> multicast reception ...
the following are true, with NET_ADMIN, NET_BROADCAST, NET_RAW,
SYS_ADMIN all set
* clustering tool's test case suggests i'm SENDING multicast
traffic, but not RECEIVING
* tcpdump suggests i'm SENDING but not RECIEVING
* i can ping the multicast address from the shell while the app
server is running, but not when its not running
* adding routes doesn't seem to effect it at all
>
> if you are interested in 'improving' multicast
> capabilities in a safe way, and willing to do
> some testing, please contact me on the IRC
> channel ...
i'm down for whatever. i have a good testbed. i just missed you on
irc tonight. will try againtomorrow.
>
> best,
> Herbert
>
> PS: will require switching to 2.6 kernel and
> recent devel version (2.1.x)
i'm planning on moving to 2.6 anyway. we can chat about the 2.1.x
stuff.
<snip>
>
_______________________________________________
Vserver mailing list
Vserver@list.linux-vserver.org
http://list.linux-vserver.org/mailman/listinfo/vserver
Received on Wed Nov 16 11:56:03 2005