On Sep 7, 2009, at 12:27 PM, Rob Lister wrote:
Hi Ralph,
Thanks for taking a look at this. First time poster - do I respond to you or logback-user@qos.ch ?
In general, when working with open source always respond to the mailing list. The mailing lists are archived and are searchable via google, etc. so others get the benefits to the answers to your questions.
I just realised that some of my text was moved to the end of my name so I didn't show the full appender;
Should have been.......
<param name="LayoutParam1" value="%-5p %x - %m%n"/> I'm struggling with this as it works fine on our internal system, but I'm not allowed to test it on our clients site (in Italy) so I can't reproduce anywhere. What diagnostic info would I need to try and put this to bed?
First, you would want to get a copy of the original log file and send it along without touching it so anyone viewing it can be sure they are seeing the original data. Zip it up and then transfer it in binary so you are sure that the data isn't being touched. Next, you might actually have to look at the code that is generating the log record in question. It may be that variables are being replaced and for some reason those variables are the source of the problem. Also, find out what code page the host system is set to and then see if you can configure your system that way. That can get tricky as your keyboard won't have the correct characters, but you should be fine for Italian since so many of the Latin 1 characters are common.
I'm a zOS sysprog so have vitually no Unix/log4j/java. It's all a bit of a culture shock at the moment.
That isn't a problem. I started out as a computer operator on IBM mainframes and was a CICS systems programmer for a few years (back in the dark ages when IBM gave you the source code). I worked at Candle Corp for 17 years working on performance monitoring products for IBM systems (MVS - z/OS, VM, AS/400) and the last several years we were doing all our work on Unix System Services so we could have a common build system across all the various operating systems we supported. I haven't worked on a mainframe since 2001 though and switched from C/C+ + to Java at the same time, but I'm not sure much has changed. Ralph