Issues with the Commodity auctions API the last months

Hi, I’ve been using the commodity auction API for a while now and to me it seems like it is barely every updating anymore.

I’ts been like this for a while. I’ve tried to ask on Discord, but I guess that isn’t the correct channel.

Would it be possible to fix this? Like right now, as far as I can tell it’s been over 10 days since the EU got an update…

What is going on here? I am using the Last-Modified header flag to check off the date.

I found the cause of the issue.

I’m getting no ‘Last-Modified’ header for the commodity auction houses for some reason.
But instead I’m getting ‘Date’.

I’m not sure if this was communicated in any way to use developers. I did not consider checking if the headers had changed…

Greetings Ukon,

This is the first I have heard about this problem. Checking the API responses, I am seeing the following timestamps. Is it possible that something on your side is normalizing the response headers or returning cached responses? Or perhaps there was a previous issue I am unaware of that has since been resolved?

US: Last-Modified: Fri, 2 Dec 2022 17:15:16 GMT
EU: Last-Modified: Fri, 2 Dec 2022 17:33:59 GMT
KR: Last-Modified: Fri, 2 Dec 2022 17:33:59 GMT
TW: Last-Modified: Fri, 2 Dec 2022 17:27:53 GMT

By any chance are you using C# with the RestSharp library ?

I ask because recently we found a case in the API discord where this lib returns the Last-Modified in a separate subset of header fields and that lead to some confusion because you get both Date and Last-Modified from the API.

Perhaps some HTTP client library you use could be doing the same ? It should always have the Last-Modified tho.

Thanks for the responses!

I found out that the actual reason was that I had provisioned too little memory. So I had to split the commodity API out into a separate lambda with more memory :slight_smile: Just forgot to update the thread with it. sorry.

I’m not sure why, but for some reason, I didn’t verify the URLs when I tested them individually to debug the issue I had. I found that in the method for manually updating one realm at a time, I forgot to authenticate, and for some reason, it did not throw an exception on 401…

I am actually using AWS Lambda with NodeJS with the Serverless framework :slight_smile:. The issue was just me failing.