For the first time since SquidFire's inception, the main screen has received a major face lift. Now the total number of rows as well as the name of the log file are listed directly from the main log list. This change will also make it easier to add additional information in the future.

Size does matter, at least when it talking about squid logs. Now requests can be searched by size and the size of each request is displayed by default in the results grid. Speaking of size, notice the number of rows in the log file being searched. This specific search scanned 100,000 rows out of 1.5 million in 11.2 seconds. Much faster than the previous version which use grep.

This is a pretty boring search, but notice how the results are listed. Actual web pages show up in bold and left justified. Secondary requests such as images and javascript requests are indented. The type of request is figured out by looking the mime type returned by the web server. However, there are a lot of broken servers out there that don't return mime types at all so the script also trys to determine the file type by the extension if it exists. A lot of work went into this feature to make it extremely accurate.

Same search as above except with the "Only Show Web Pages" option. Notice that 30,117 rows were searched this time before the maximum of 1,000 results was reached. Without this option checked only 4,826 rows were scanned before the max result limit was reached.

I left the best screen shot for last. This is a search through my largest test log file for the word "Mississippi". My 2GHZ development box finishes this search in only 148 seconds, which I think is pretty good. First, this file is compressed so it has to be uncompressed on the fly as its searched. The file is about 200mb uncompressed and has more than 1.5 million rows of requests. This same box can simply read and parse this same file in 96 Seconds so there is about 50 seconds of overhead doing the actual search as well as mime-type detection.