On 17/09/2008, at 11:35 AM, Reza Razavipour wrote:
Yes, this is a great tool. What I need to do is to do this programatically. So, the application, at any point in time, will have targets that have log4j log files. At user's request, which would be a get of all files, with some date/time constraints, the application will have to combine the log files, filter on date/time and present the log entries in "some tbd" format... What I was wondering is that: 1. Does log4j has anything to support such things as concatenation of log files, searching and filtering of log files? 2. If not, is there any other package that does similar things.
The pinpoint-importer layer can import a log file into an indexed 'context' area. You could then point (pardon the pun) the pinpoint- search layer at that generated context and do programatic queries using Lucene (date ranges, free text etc) to get a PinpointSearchResults. You can then iterate over the results to get the raw LoggingEvents that have matched to then export to some format of your choosing.
I'm planning on adding a class that converts the search results->Excel via the POI library at some point. My main focus has been trying to design the query mechanism to be able to scan quickly for production triage purposes.
I'm not exactly sure what your use cases are, but I would have thought a background mechanism to soak the log files into a Pinpoint context so that at the point the user needs to search you have all the data there ready to go would be better.
My goal is to produce a Web application that allows an Engineer or QA person to search logs for problems based on User reports, find their logs, and work out what they were doing. It's quite a difficult task when one has Gb's worth of logs to find specific information in a sea of logs (yes there's a lot that can be done with grep/sed/awk, but a data-mining tool is much more friendly..)
cheers, Paul --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
