[
https://issues.jboss.org/browse/JGRP-2210?page=com.atlassian.jira.plugin....
]
Bela Ban commented on JGRP-2210:
--------------------------------
I agree with Bruce's conclusion: when JGroups tries to read the 2-byte version number,
it gets an EOFException, because the packet's length is less than 2 bytes (1 or 0).
I will add some code that checks the length of the received message (or message batch) and
drops it if the length is less than N bytes. I need to look into what N should be but I
guess 3: version (2 bytes) plus flags (1 byte).
While this causes an exception, I assume it does not cause incorrect behavior, does it?
EOFException[JGRP000030] : failed handling incoming message
-----------------------------------------------------------
Key: JGRP-2210
URL:
https://issues.jboss.org/browse/JGRP-2210
Project: JGroups
Issue Type: Bug
Affects Versions: 3.6.10
Reporter: Dharam Thacker
Assignee: Bela Ban
Hi Team,
I am using Apache geode as our distributed caching solution which internally using
JGroups 3.6.10.Final version.
I see this exception almost daily 5-6 times but we don't know the root cause for
that. I tried reaching with Apache Geode community as well. As per their suggestion, it
could be due to empty datagrams being sent to process.
I confirm that both of my client and server processes are using same version as
3.6.10.Final.
I would request you to look into the same.
*+Exception Stack Trace:+*
[error 2017/07/28 00:13:30.521 EDT EventServer <unicast receiver,hostXXX>
tid=0x45]
JGRP000030: hostXXX<v2>:1025: failed handling incoming message:
java.io.EOFException
java.io.EOFException
at
org.jgroups.util.ByteArrayDataInputStream.readShort(ByteArrayDataInputStream.java:138)
at org.jgroups.protocols.TP.handleSingleMessage(TP.java:1705)
at org.jgroups.protocols.TP.receive(TP.java:1654)
at
org.apache.geode.distributed.internal.membership.gms.messenger.Transport.receive(Transport.java:160)
at org.jgroups.protocols.UDP$PacketReceiver.run(UDP.java:701)
at java.lang.Thread.run(Thread.java:745)
Within our company, there is some cyber security server which keeps hitting our process
& I get below warning as well,
[warning 2017/08/16 22:05:51.334 EDT EventServer <unicast receiver,hostXXX>
tid=0x46] JGRP000010: packet from *hostYYY*:45482 has different version (0.0.2) than ours
(3.6.10); packet is discarded
*hostYYY* is not a part of my cluster process [Neither client not server] but some cyber
security scanner which knows about all hosts and have its own way to check for security
violations
Response from Geode community:
https://www.google.co.in/url?sa=t&rct=j&q=&esrc=s&source=...
--
This message was sent by Atlassian JIRA
(v7.2.3#72005)