问题
My application allows users to "link" their YouTube accounts to our system and then we allow them to upload videos to their channels both automatically in some cases, and on an "upload" action on their part. We have hundreds of videos being uploaded because we have thousands of users that use our system.
Today I saw hundreds of errors in my application when our batch automatic upload job was running. The errors were for:
quotaExceeded, video upload limit reached.
My API quotas are very high (50,000,000 per day, 3,000,000 per 100 seconds, 300,000 per 100 seconds per user), so if a video is 1600 points, I have enough limit to upload 30,000+ per day, 1,800 videos per 100 seconds, and 180+ per 100 seconds per user.
I have seen other questions out there hinting at some newly enforced limit by Google to just 50 videos and then 1 video every 15 minutes? This is a very low limit for my application which has such a large quota! To be clear, at most I have seen my application handle 1,000 videos in a single day (1,600,000 points of my total 50,000,000 quota).
Is there any way I can get this artificial limit of 50 videos and then 1 per 15 minutes removed? This is a major block to my users' functionality for a very popular web app. I could understand if it were 50 videos per user and then 1 every 15 minutes, but I highly doubt my errors this morning were from one user trying to upload 300+ videos at a time. My system only uploads their "newest" videos each day, which most people only have 1-10 videos at max. Hundreds would be a very rare edge case.
回答1:
You can actually confirm in your Developers Console the available quota for your application. By default, YouTube API have a quota allocation of 1 million units per day as mentioned in Quota usage. If you see that your usage reached your quota limit, you can request additional quota on the Quotas tab.
Note also that, all API requests, including invalid requests, incur a quota cost of at least one point. You may use the Quota Calculator to get an estimate of the quota cost for an API query.
On the other hand, to work efficiently with your quota and if you haven't done so, I suggest that you implement exponential backoff if you're encountering high error ratio. See this sample code which shows an exponential backoff strategy to resume a failed upload. Also, if applicable, subscribe to Push Notifications which is much more efficient than polling-based solutions.
Check the documentation for more information on how PubSubHubbub callback server receives Atom feed notifications when a channel does any of the following activities:
- uploads a video
- updates a video's title
- updates a video's description
Hope that helps!
来源：https://stackoverflow.com/questions/43158150/our-application-uploads-hundreds-of-videos-per-day-using-the-youtube-api-but-no