How should I store documents for a logging application? - ravendb

Question

I am writing a logging application and I'm using RavenDB as the data store. We have a lot of textual based logging and want to be able to leverage the indexing abilities of RavenDB to search through it.
I started writing my document implementation like this:
public class LogEntry
{
public DateTime Date {get;set;}
public string Message {get;set;}
}
public class Information : LogEntry {}
public class Error : LogEntry {}
I did this so the UI would show different collections for different log levels, however, after thinking about it some more, I would only be using the UI to query the logs.
Since RavenDB stores all of the documents the same behind the scenes, would it be better to just have a single document LogEntry that exposes a severity level property, and then create multiple indexes that group document collections by severity level?
public class LogEntry
{
public DateTime Date {get;set;}
public string Message {get;set;}
public string Severity {get;set;}
}

Answer 1

I would create a single class, but for Severity, how about an enum type? RavenDB will store enums as strings, so it doesn't have any performance impact. But you'll get the benefit of avoiding typos and having Intellisense in your .Net code.
You should just have one index, not multiples. Your map should include the any fields that you want to sort or filter on, so it may be that you map all three fields, Date, Message and Severity.
You might also want to consider marking the Message field as "analyzed" in your index. This will allow full-text searches of your log messages.
public enum Severity
{
Debug,
Info,
Warning,
Error
}
public class LogEntry
{
public DateTime Date { get; set; }
public string Message { get; set; }
public Severity Severity { get; set; }
}
public class LogEntriesIndex : AbstractIndexCreationTask<LogEntry>
{
public LogEntriesIndex()
{
Map = entries => from entry in entries
select new
{
entry.Date,
entry.Message,
entry.Severity
}
Index(x => x.Message, FieldIndexing.Analyzed);
}
}