• Hi there,

    Since this morning this seems to be happening;

    Better Font Awesome
    
    API Error
    The attempt to reach the jsDelivr API server failed with the following error: http_request_failed: cURL error 28: Operation timed out after 10001 milliseconds with 0 out of -1 bytes received (URL: https://data.jsdelivr.com/v1/package/gh/FortAwesome/Font-Awesome)
    
    Don't worry! Better Font Awesome will still render using the included fallback version: 4.7.0
    
    Solution
    This may be the result of a temporary server or connectivity issue which will resolve shortly. However if the problem persists please file a support ticket on the plugin forum, citing the errors listed above.

    It this something we can circumvent?
    It’s causing slow loading on backend / frontend now.

    FYI; for now I disabled the plugin.

    Thanks in advance for your reply.

    Kind regards,
    Patrick Dankers

    • This topic was modified 6 years, 4 months ago by Patrick_D1985.
Viewing 12 replies - 1 through 12 (of 12 total)
  • I had this too:

    API Error
    The attempt to reach the jsDelivr API server failed with the following error: http_request_failed: cURL error 28: Operation timed out after 10001 milliseconds with 0 bytes received (URL: https://data.jsdelivr.com/v1/package/gh/FortAwesome/Font-Awesome)

    Thread Starter Patrick_D1985

    (@patrick_d1985)

    I ended up disabling it for now.
    We have a custom fallback in place which provides FA in another way … but does not offer the icon selection like BFA does in the backend.

    Hope this gets resolved as it hurts more then it should in performance really.
    Not saying that it’s the blame of the author btw, but I think it should be a shorter check. The files are either available or not that should not even need 10001 milliseconds to determine.

    I have not disabled it, but the pages are taking a really long time to show. I have no alternative in place so don’t know what to do.

    same here:

    API Error
    The attempt to reach the jsDelivr API server failed with the following error: 503: Service Unavailable (URL: https://data.jsdelivr.com/v1/package/gh/FortAwesome/Font-Awesome)

    Hello, this is an issue with JSDelivr, if you want a quick fix (at least for the horrendous loading times) I suggest you change the timeout (It’s at 10 seconds..) in the files.

    The file is in the path PLUGINS\better-font-awesome\vendor\mickey-kay\better-font-awesome-library and the file itself is better-font-awesome-library.php

    Edit using a program called Notepad++ if you FTP (and not default Notepad) or using the command line editor like Nano in Bash.

    The Timeout function is located at row 115, the string will look like:

    	private $wp_remote_get_args = array(
    		'timeout'   => 10,
    		'sslverify' => false,
    	);

    Change 'timeout' => 10, to whatever you desire, I personally changed this to 'timeout' => 1, and saw the site go way faster. This should be a fine intermeditary fix and everything should function as they used to when JSDelivr is back up and running.

    Sincerely,
    Simon

    Plugin Author Mickey Kay

    (@mcguive7)

    Hi folks, I reached out to jsdelivr, and they confirmed there is currently an issue with their API unfortunately. See: Check out @jsDelivr’s Tweet: https://twitter.com/jsDelivr/status/1058820333745119233?s=09

    The good news is at the front end of your site should not be affected, however I’m so sorry for the inconvenience. I will be following up with them to ascertain what the resolution timeline is here.

    Thank you for checking with JsDelivr, however I don’t think a 10 second timeout is reasonable as it slows down every action (even changing page in admin area) by 10 seconds

    This should only be a beta issue though (I hope) and using beta plugins on production websites isn’t optimal

    Thread Starter Patrick_D1985

    (@patrick_d1985)

    Thanks for the feedback to both of the above.

    Mickey is there a reason for the 10s timeout?
    You are relying on a third party here and I think having it lower by default would be better as to ensure a workable situation in the long run.

    As mentioned above … it slows every thing in the backend.
    Either way thanks for looking into this!

    As for beta running on production; agreed but I hope that a FA5 supporting version comes out of beta sooner then later ??

    • This reply was modified 6 years, 4 months ago by Patrick_D1985. Reason: mention did not go as planned

    Hey all, I am from the jsDelivr team. I am really sorry about the issues our API had.
    It was related to our Redis instance running out of RAM. Only our API was impacted, and only partially for un-cached URLs. So the outage was fairly limited. The CDN itself had no issues at all!

    I wanted to let you know what we plan to take precautions and avoid this kind of problems in the future. We take this kind of issues very seriously and plan to make our API backend even more reliable.

    But I also wanted to let you know that while our CDN at cdn.jsdelivr.net has multiple failover strategies and uses multiple providers, our API does not share the same infrastructure. Ideally applications using it should have a failover logic to ensure an API timeout does not break anything. For example cache API responses and re-use the cache if the API is having issues. This applies for all APIs in general.

    Regardless we will be making sure you nothing similar will happen ever again!

    I am really sorry for the inconvenience this caused everyone

    Thank you for clarifying the issue Jimaek

    Thanks for fixing it quickly for us.

    Yes, thank you!

Viewing 12 replies - 1 through 12 (of 12 total)
  • The topic ‘JSDeliver issue’ is closed to new replies.