IndexNow Timing Out

[Emediate]

I'm getting this System Log a couple of times a day:

IPS\Http\Request\CurlException::28

https://api.indexnow.org/indexnow/
Operation timed out after 10000 milliseconds with 0 bytes received

Don't think there's anything I can do on my end?

[Marc]

You would need to check the connectivity between yourself and indexnow to determine where the timeout there is. As this is an issue on the connectivity side of things, there is nothing we can do from a software point of view that would prevent this

[Emediate]

That's all I need to know. Will look further down the line now, thanks 🙂

[Marc]

1 hour ago, Emediate said:

That's all I need to know. Will look further down the line now, thanks 🙂

No problem. Good luck in getting the issue resolved

[Emediate]

On 4/6/2022 at 10:18 PM, Marc Stridgen said:

You would need to check the connectivity between yourself and indexnow to determine where the timeout there is. As this is an issue on the connectivity side of things, there is nothing we can do from a software point of view that would prevent this

Just a follow up, our hosts have confirmed they can ping the IndexNow and GoogleAPI (also timing out) from our server, and from the timestamps in the IPS Error Logs, there is nothing that corresponds with the server logs. They believe the issue still relates to Invision.

Anyone else come across timeout errors to IndexNow?

[Thomas P]

4 hours ago, Emediate said:

Anyone else come across timeout errors to IndexNow?

We have this error every once in a while.

How big is your forum, i.e. how many posts do you carry or how many new posts are created during one day?

[Emediate]

4.5M+ posts and yes hundreds of posts per day. 

[Emediate]

Problem now out of sight. IndexNow disabled. 

Will revisit later when the bug is probably realised. 

[Stuart Silvester]

37 minutes ago, Emediate said:

Problem now out of sight. IndexNow disabled. 

Will revisit later when the bug is probably realised. 

This isn't a bug in our software, it's an issue between cURL and the remote service. There may be something intermittently blocking the request for some reason.

[Emediate]

Nevermind.

Edited April 11, 2022 by Emediate

[Stuart Silvester]

Quote

Yep. But experience now tells me to wait a bit until other people report it and then it gets looked at and patched.

IndexNow is still referencing a file that is supposed to be in forum root, that doesn't exist.

But either way, I've disabled this now and moved on.

We released a patch for that issue, if you're still seeing it then that's a problem we need to look into (although I am unable to reproduce the problem). However, that isn't what you're reporting here.

Timeouts are an issue between the network and the remote server, it hasn't responded in a pre-allocated amount of time.

[Emediate]

This has gone away since disabling Indexnow (obviously), but I still get daily time outs to:

Quote

https://fcm.googleapis.com/fcm/send/e_XMYen5e2M:APA91bH8ar2pMEZltgB_ZBzE5hoHiylyesetAGOkqMwriVz0bQI_gTU1NSmjLkfllNlZL-ffUlqXIxM31mdWe2Wdt7rG_fZ1eiSJhu4xV-rz6N7Isy5XKXnf2m86cm7vNizUX1LKPxS-
Connection timed out after 10000 milliseconds

Quote

https://fcm.googleapis.com/fcm/send/fBSr8Xxg448:APA91bFxaG3y2ekhy7jB0XYQzS2eiOfP611KkJzfmoIHFX484-X5re-sd2jZcTf2fA_9efM22bq5JlHWqgbcsY7G0Fvy0u6j-nbYUx-eeWb-PHZU3RSttk9SvvnAy6ak647EQjaurBx6
Connection timed out after 10001 milliseconds

Quote

https://fcm.googleapis.com/fcm/send/dpU_ONaPQNI:APA91bE1OGYGVDIBLVhQaUboHedtvYLhwjs4Xpg4z-BjEsKP6Vr-Rh90_NkhN-ElsIhgnETMRUGjM98Rmsi1mAZoBRRTaqi3mPMAn7fElLej9DyXUwAnZwtGgGCrt6goAm-2nt0nuv4-
Connection timed out after 10001 milliseconds

Quote

https://wns2-sg2p.notify.windows.com/w/?token=BQYAAAA0KSyfDWZjWUuBKWUkDoWWA4YTxayCiRKL%2fvSqU6riPmEHH60FhlodkTswcsbwcalkdKAT8iVweAJH3WeHEulRvR0mCBdBeUH%2bMZ0Ihy2p15tKchEX1KBcmaN%2bJizu4pEDpqUM%2fLBzUcJounNCqmlEC6E%2boSM7DXjJLDSbu5PpJY5TPGdT5JFGYrJPb2s50ajPLuDMrvVedMzeG8li%2b40xGfBjtmptDQ9Cp%2bqyFrkC7rhCqvw0Gfsmp0kFQwjBOUfyddOheUu8igrkTXxKKA3lxWtIGGcWiPEp4ixXAnmPLUPbV4TCrz%2fG5Z%2baelASOUA%3d
Connection timed out after 10001 milliseconds

Quote

https://www.google.com/recaptcha/api/siteverify
Connection timed out after 10003 milliseconds

Our hosts have checked these logs and timestamps against the server logs and there is nothing that corresponds there. The hosts have also pinged each of those with no issue confirming nothing in the firewall blocking these connections. But they do time out daily from the IPS software.

Any ideas or tips please?

[Randy Calvert]

I've seen this specifically on the IndexNow service, but I monitoring in place to monitor IndexNow and in those cases, it was not my server, but the route in between my server and IndexNow.  

So this is most likely not a problem on your server itself, but somewhere on the internet in between you and the destination server.  The internet is like the roads/highway system.  

You leave your house and mid-way through your trip, it's possible you run into traffic.  It's not a problem on your source or destination side that made you late, but instead something along the way.  This is saying there was no response within 10 seconds of the request so it gave up waiting.  It does not mean your server was blocking it, but that there was a temporary issue that typically goes away within a few minutes.   

[Marc]

As mentioned, this is a connection issue somewhere between yourself and their server at those points in time. While I understand your hosting company has checked this, if its intermittent, they may indeed see it working. Given its different services though, I would say the issue is more likely on your server side than anywhere else. The software itself cannot cause a timeout.

[Emediate]

Surely I am not the only one getting these repeated and daily Error logs then?

[Randy Calvert]

1 hour ago, Emediate said:

Surely I am not the only one getting these repeated and daily Error logs then?

It does not happen DAILY for me, but it’s not an irregular occurrence.  At the end of the day, it’s not a problem with IPB itself.  It’s a problem between your server and IndexNow.  IPB can’t fix that however.  

You could setup some sort of monitoring from your server on a per minute basis to monitor the API site for IndexNow.  However that’s outside of the support that can be provided here.  Once you have the info, you would have it to work with your host and show them proof of the recurring issue.  

Literally that error means IPS sent the request and got nothing back after 10 seconds.  There’s not much more they can do.  You don’t want to hold requests for tremendously long times.  It hurts server load and causes more problems than just some error log messages.  

I personally just ignore/delete them.  At the end of the day, 95% of my traffic does not come from Bing/Yandex.  It’s coming from Google and Google does not use it.  So unless Bing or Yandex is a significant source of traffic for your site, you’re most likely better just turning it off.  It has little “real world” value if Bing/Yandex is not where the lion share of your users are getting to you from.  

Edited April 20, 2022 by Randy Calvert

[Marc]

Randy pretty much has it covered there. From a software standpoint, we have no control over that unfortunately

[Emediate]

Ok im letting it go now. 

[Square Wheels]

I have a VPN with three sites.

Only the busy site is showing errors.

If this were a server error, wouldn't I get errors in all three sites System logs?

[Randy Calvert]

No. It could be the index now server rate limiting you, it could be a transient problem between requests, etc. 

Literally that error says IPB sent a request to IndexNow and did not hear back in 10 seconds. There is nothing else that can be done. It literally CANT be a code problem unless that just change it to wait longer (say 30 seconds). But if it does not finish after waiting 10 sec is it going to really request if you wait longer?