APAR status
Closed as program error.
Error description
When replicating an object using DRS/Disthub, MAX_MESSAGE_SIZE
(parameter introduced by
PQ91090) can be exceeded and then the
broken pipe and EOFExceptions cause a connection reset and
subsequent exceptions.
Change the behavior so that when a message is encountered
that is larger than the MAX_MESSAGE_SIZE, it is discarded rather
than doing a connection reset. This change will prevent
subsequent exceptions and reset attempts.
Local fix
1) do not exceed the MAX_MESSAGE_SIZE.
2) Increase the MAX_MESSAGE_SIZE to largest object's size.
Problem summary
****************************************************************
* USERS AFFECTED: All users of the Websphere Application *
* Server using HTTP session memory to memory *
* replication or Dynamic cache replication. *
****************************************************************
* PROBLEM DESCRIPTION: Users of the Data replication service *
* observe frequent resets of connections *
* to their replicator. *
****************************************************************
* RECOMMENDATION: *
****************************************************************
By default, the data replication service component can
replicate objects that are of the order of 2MB. This means
that if http sessions of size larger than this are replicated,
or if the dynamic cache service attempts to replicate cached
objects larger than this size then the replicator resets the
connection associated with such a data replication service
client. The use of a fix older than this one allows you to
customize the value of this default max_message_size. This can
be done by setting a system property by name MAX_MESSAGE_SIZE
to the value that you would like to. The units are in bytes.
This fix also allows you to set a boolean system property
named DISCARD_LARGE_MESSAGES to true, whereby the Replication
service will compute the size of any message that is being
replicated. If its size exceeds the value of the configured
MAX_MESSAGE_SIZE, then such a message will be discarded and
not replicated. This eliminates the cause of the connection
termination and hence the reset.
It should be noted however that for best performance, the
size of the replicated objects must be kept under 2MB.
The method to set the System property is as follows:
Click on the following in the admin console:
Servers -> Application Servers -> <your server> -> Process
Definition -> Java Virtual Machine -> Custom
Properties -> New.
Create a new custom property with
NAME VALUE
------- -------
DISCARD_LARGE_MESSAGES true.
Problem conclusion
This fix will be available as part of 5.0.2. cumulative fix
7 (5.0.2.7) and 5.1.0 Cumulative fix 6. (5.1.0.6).
Temporary fix Comments
APAR information |
APAR number |
PQ91650 |
Reported component name |
WAS BASE 5.0 |
Reported component ID |
5630A3600 |
Reported release |
00S |
Status |
CLOSED PER |
PE |
NoPE |
HIPER |
NoHIPER |
Special Attention |
NoSpecatt |
Submitted date |
2004-07-19 |
Closed date |
2004-08-02 |
Last modified date |
2005-01-18 |
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
Modules/Macros
Publications Referenced
Applicable component levels |
R003 PSY |
UP |
R00A PSY |
UP |
R00H PSY |
UP |
R00I PSY |
UP |
R00P PSY |
UP |
R00S PSY |
UP |
R00W PSY |
UP |
R103 PSY |
UP |
R10A PSY |
UP |
R10H PSY |
UP |
R10I PSY |
UP |
R10P PSY |
UP |
R10S PSY |
UP |
R10W PSY |
UP |
|