HTTP Status Code 499

[joniveehernandez]

Hi guys, just within this week we've encountered a problem on our objects using HTTP. It suddenly having 499 problem debug the scripts and php code to see what is the problem but nothing goes around. Tested the response status in LSL it seems like its having hard time to get a response from the outside server. Tested it few times and response sequence goes like this: 200, 200, 200, 499, 200, 200, 499, 499. Another thing those objects are used for almost a year now without problem and never been modified and updated by the timeline. And very awkward that my team mate object that he developed before is also having the same problem AND to clarify the we are not using the same server. Is this something to do with SL itself or problem between SL and Outside Server?

[Ron Khondji]

This is from the wiki:

 

• If there is a space in url, the http_response status code will be 499.
• Requests must fully complete after 60 seconds, or else the response will be thrown away and the http_response status code will be 499.

 

I'm guessing the problem is either with your server or with the connection between SL and the server. 

 

[Freya Mokusei]

200 is a standard HTTP response code which means OK, which of course means the server is responding and happy to reply.

499 is one of Second Life's proprietary (non-standard) HTTP error codes - it means several things, like Ron has already said. In addition to those two conditions, there's some others:-

• It can mean that you're sending requests too quickly - from memory you have a limit of 100 http-requests per 100 seconds, but this may have changed
• It can mean that the simulator is too busy or cannot allocate a request for technical reasons
• SSL failure

If the 200 codes (which succeed) aren't performing requests/postbacks as you'd expect then your problem isn't the 499 code directly, you might just be sending requests too quickly.

[joniveehernandez]

Does it mean I'm sending multiple requests to the server in all the multiple objects across all our regions? But we've been using this fine for almost a year and we encountered this problem just last within the week.

[Freya Mokusei]

As I have no familiarity with the system in question I can't say. It is possible that your server isn't answering requests quickly enough. It could also depend on how many of these objects you have per region, and how often they send requests. It is suspicious that it only started failing recently, if any of these settings can be varied globally then it might be worth experimenting to see if you can improve reliability.

Have you checked the HTTP logfiles on your server? Does this provide any clues by analysing requests against the sc-status, sc-bytes, cs-bytes or time-taken columns in your logs?

Does the system 'work' during 200 codes? Can it not simply bridge across the 499 codes, or queue them and attempt to retry?

[Levio Serenity]

Random 499s are a part of life in SL scripting.  All you can do is watch for them and if you get one, wait awhile and resend the request.  Of course your server has to be able to gracefully handle getting a duplicate request since it may be sent multiple times. Sometimes customers will report a lot of 499s from a single sim, in that case, ask them to have the sim restarted.