Yes, parsing is the biggest issue. There are all kinds of possibilities with the format. For example they could use a null terminator for the records rather than a new line terminator. That is why I thought allowing list files in the jboss.server.log.dir directory was the best option.

Essentially there would probably be two operations. One that lists the available files and one that reads the contents of the file. The read would probably be best in chunks since big files could be an issue. The main security issue here would be if someone starts the server with -Djboss.server.log.dir=$JBOSS_HOME/standalone/configuration they would have access to files like the This is one of the reasons I thought maybe it should be required to have a .log extension.

One idea I just thought of, and it might not even be possible, but it maybe have a way to email a compressed version of the log file. I could see some benefit with large files, but maybe ssh is better for that.

On 08/15/2013 08:14 PM, Jason Greene wrote:
I forgot to mention that I chatted with David late last night and he pointed out cases where applications dump multiline data that makes parsing records impossible.

He also brought up a problem with expression based file names, that could make it not possible to reverse rolling logs.

We did come up with several potential solutions but we really need to nail the use cases first.

The download/view case could be done securely and address the rolling expression problem by making the feature more of a file browser that is locked into a set of directories. By default we would allow browsing into the log directories. 

On Aug 15, 2013, at 6:45 PM, "James R. Perkins" <> wrote:

On 08/15/2013 03:27 PM, Harald Pehl wrote:
I agree that we should focus on the most important use cases:
- Not able / not like to sshing into the server to look at the logs
- Deployment failed and you want to know why (staying in the console)

As for the console I would suggest to start with a very limited set of features. No filtering, just show the last n log entries, where n is specified as part of the  management operation. At first the user would see the last n log entries and can navigate further backwards in n-steps. For all deeper analysis the user should be able to download the full log. When showing the log to the user, I can think of some kind of syntax highlighting (see for an example).
The problem here is we don't know the last n entries if we read a file. We can read back the last n bytes or possibly lines, but that could contain portions of a stack trace.

That said there are however use cases where parsing / filtering makes sense. The audit log has a well defined format ( and in the current implementation I'm using a master / detail view to display the audit log items.
This might be a little easier for audit-logging since we have more control over what's happening. I have to look at the code, but we might just want to always write a formatted file in addition to other types like the syslog.

.: Harald

Harald Pehl
JBoss by Red Hat

Am 15.08.2013 um 03:11 schrieb

On 8/14/2013 5:20 PM, Brian Stansberry wrote:
Agreed. IMHO this is the most important driver for this feature.

On 8/14/13 3:44 PM, James R. Perkins wrote:
Just to add too in a domain it would be nice to have a central spot to
view logs instead of having to ssh into various servers.
Also realize that in many shops you don't have the option to ssh into
various servers.  If they are serious about security then access to the
console is likely to be the only access you have.

On 08/14/2013 12:56 PM, Jason Greene wrote:
Mainly convenience. You deploy something it fails, you want to look at the log but don't feel like sshing into the server. As to performance the cost would be on request, and not more expensive then looking at the log file via ssh.

On Aug 14, 2013, at 2:24 PM, Scott Marlow <> wrote:

What are the use cases for online reading of the server logs?  If there
are problems occurring on the application server (e.g. perhaps the cpu
is pegged), reading logs online, could make the system even less

If we just want to read the server logs as part of a health check, not
requiring the server console to be working would be better.

Should the reading of the logs instead be an external capability?
Perhaps using the logs from the JBoss/WildFly Diagnostic Reporter output
(archive) or some other archived copy of logs.

Another compromise, add the WildFly Diagnostic Reporter (or at least the
log collection part) to the management console (output archive is
downloaded for local viewing).

On 08/14/2013 01:03 PM, James R. Perkins wrote:
I had posted this to another list, but this is a more appropriate place
for it. I think there needs to be a general discussion around this as
it's been mentioned, at least to me, a few times here and there and I
know Heiko raised the issue some time a go now.

The original JIRA, WFLY-280[1], is to display the last 10 error messages
only. To be honest I wouldn't find that very useful. To me if I'm
looking for logs I want to see all logs, but that's not always so easy.
Like the syslog-handler which doesn't log to a file so there is no way
to read those messages back.

The current plan for the last 10 error messages is we store messages in
a queue that can be accessed via an operation. This works fine until the
error message you're interested in is 11 or you want to see warning

Another option I had come up with is reading back the contents of the
file, for example the server.log. This could be problematic too in that
there is no way to filter information like only see error messages or
only see warning messages. To solve this I have considered creating a
JSON formatter so the results could be queried, but I don't think it
should be a default which would mean it's not reliable for the console
to assume it's getting back JSON.

I've also thought about, haven't tested this and it may not work at all,
creating a handler that uses websockets to send messages. I'm not sure
how well this would work and it's possible it may not even work for
bootstrap logging.

With regards to audit logging, we're probably going to have to do
something totally different from what we'll do in the logging subsystem
since it doesn't use standard logging.

I guess the bottom line is what does the console want to see? Do you
want to see all raw text log messages? Do you want all messages but in a
format like JSON that you can query/filter? Do you really want only the
last 10 error messages only? All or none of these might be possible, but
I really need to understand the needs before I can explore more in depth
what the best option would be.


James R. Perkins
Red Hat JBoss Middleware

wildfly-dev mailing list

wildfly-dev mailing list
Jason T. Greene
WildFly Lead / JBoss EAP Platform Architect
JBoss, a division of Red Hat

wildfly-dev mailing list

wildfly-dev mailing list

James R. Perkins
Red Hat JBoss Middleware
wildfly-dev mailing list

James R. Perkins
Red Hat JBoss Middleware