Log file reliability

[AndyP]

A customer told me an employee submitted a form on their website at a specific time and they said it didn't go through. From the "Raw Access Log" I can see that no one accessed the site within two hours of when they said they submitted the form. So I have two questions:

1. Is there any condition in which lines would be added to the log out of chonological order (which still doesn't matter because I searched the whole thing)?
2. How reliable is this log and are visits ever omitted for any (or no) reason?

[AndrewT]

As far as HTTP logs your answers are No and never, they are very reliable. It easily could have been an issue with the submitting user's internet connection which ended up not requesting the page to begin with. The only exception to this is if you access something via the preview URL, it will log it elsewhere.

[Frank Hagan]

A customer told me an employee submitted a form on their website at a specific time and they said it didn't go through. From the "Raw Access Log" I can see that no one accessed the site within two hours of when they said they submitted the form. So I have two questions:

1. Is there any condition in which lines would be added to the log out of chonological order (which still doesn't matter because I searched the whole thing)?
2. How reliable is this log and are visits ever omitted for any (or no) reason?

Silly question, but did you allow for the server time being different from the user's time? Quite a few scripts show the user their own time zone, adjusting the server time plus or minus the number of hours to their time zone. I've had more than one customer be confused by this, as the "website" is telling them the time they have on their own computer, while the server itself is really operating on US Central Time.