Re: Rmi Client Crash

On Monday 28 March 2005 16:55, gb c wrote:
> When server calls a method of crashed
> client, server waits for minutes then other clients
> give timeout errors

I think you are describing something different.
Servers don't call methods on clients.

So we need to know what you are trying to say.

The general statement; "My problem is that when a client crashes,
system crashes." is not an RMI problem, and most likely your coding style
(e.g. not catch exceptions).

General Info;
The RemoteServer can not know if a client has disappeared, for instance, if
you switch off the power the connection is not terminated at the server end.
Not until some communication over the connection occur can the condition be
detected.
However, the opposite is different. When you call a RemoteServer and the
connection has not been established, a connection is made after which the
remote call occurs. If there is a connection problem, the calling client will
know at that point. Long delays in finding out that the server can not be
reached, is typically due to improper network setups, where typically Windows
clients will timeout over various Windows name resolution systems before
timing out completely. This can take a minute or more.

The RemoteServer has a lease system for the remote references, and if the
client doesn't update the lease, the remote reference is garabage collected
and potentially the connection is closed.


Cheers
Niclas

===========================================================================
To unsubscribe, send email to [email protected] and include in the body
of the message "signoff RMI-USERS".  For general help, send email to
[email protected] and include in the body of the message "help".

For a list of frequently asked RMI questions please refer to:
http://java.sun.com/j2se/1.3/docs/guide/rmi/faq.html

To view past RMI-USERS postings, please see:
http://archives.java.sun.com/archives/rmi-users.html