Error polling uplinks: Unexpected token '<', " - When polling all uplinks

[ecasey]

This seems to almost randomly occurs, but keeps coming back. Seems to imply maybe a bad json or other config file ; unless it
is tokenizing something else during the poll.

But it keeps coming back. I'm using 1.8.2, but have seen it in earlier versions also. Have never been able to determine the source of the problem.

[awehttam]

This happens when the api end point isn't returning a proper response or if there's a proxy error. Check:

1. The Apache error log for the site for any errors around the same time
2. F12 developer console, see if there are any script errors. The network inspector can also show the exact response if it's open when the error occurs

I don't see this error on my systems.

[ecasey]

reviewing network traffic saw gateway timeouts were occurring and increased connections from 10 to 20 helped; but have seen it reoccur once more. still observing and trying to determine cause.

[ecasey]

must be something changing in my config . it cleared up after fixing spaces scheduling definitions.

But it is back :-( invaild token "<" and gateway time out....

[awehttam]

I'm curious if this gets better with 1.8.5. I made some changes to how the admin_daemon works because certain elements of the site would back log when waiting for a reply (like if it was polling in the background, the admin interface wouldn't load data until the poll completed).

It may also be the length of time the polling goes on. How long does it take before the timeout occurs?

[ecasey]

Looks to be 70-75 seconds. Remember I have a lot of networks. Addresses:fidonet1:3634/58.1337 lovlynet227:1/11 hobbynet954:100/66.1337 tqwnet1337:3/194.1337 fsxnet21:1/182.1337 pinet314:413/21.1337 cnet64:500/21.1337 retronet80:774/21.1337 weednet420:1/16.1337 spooknet700:100/78.1337 health999:1/66.1337 fishingnet701:1/25.1337 wwivftn11:1/238.1337 ilink454:3/158.1337 nixnet86:100/12.1337 whispernet316:36/39.1337 zeronet911:1910/0.1337 metronet25:25/40.1337 disciplenet12:316/21.1337familynet63:10/104.1337

…

-- Errol Casey ***@***.*** Read some Messages, Play Games, or Chat! Try my BBS - https://toolazy.synchro.net/bbs

On Monday, March 2nd, 2026 at 4:43 PM, Matthew Asham ***@***.***> wrote: awehttam left a comment [(awehttam/binkterm-php#156)](#156 (comment)) I'm curious if this gets better with 1.8.5. I made some changes to how the admin_daemon works because certain elements of the site would back log when waiting for a reply (like if it was polling in the background, the admin interface wouldn't load data until the poll completed). It may also be the length of time the polling goes on. How long does it take before the timeout occurs? — Reply to this email directly, [view it on GitHub](#156 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AB5JJBSVKDXUGG7YEU3Q4JL4OX6BJAVCNFSM6AAAAACWA3NUMCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTSOBXGA3TCMRZHE). You are receiving this because you authored the thread.Message ID: ***@***.***>

[awehttam]

It sounds like a timeout to me, either PHP's maximum execution time or the proxy has a maximum time it waits for a response. If you're using PHP FPM there could also be another timeout occurring. The only thing I can do is change it so that the poll all command returns immediately, which means you would not see any response output from the poll itself.

[ecasey]

immediately returning without output doesn't sound good. maybe I can tune the php timeouts on my end ? i'll do some digging.

…

-- Errol Casey ***@***.*** Read some Messages, Play Games, or Chat! Try my BBS - https://toolazy.synchro.net/bbs

On Tuesday, March 3rd, 2026 at 6:42 PM, Matthew Asham ***@***.***> wrote: awehttam left a comment [(awehttam/binkterm-php#156)](#156 (comment)) It sounds like a timeout to me, either PHP's maximum execution time or the proxy has a maximum time it waits for a response. If you're using PHP FPM there could also be another timeout occurring. The only thing I can do is change it so that the poll all command returns immediately, which means you would not see any response output from the poll itself. — Reply to this email directly, [view it on GitHub](#156 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AB5JJBUSLVQKWPMXWICFMIL4O5UXZAVCNFSM6AAAAACWA3NUMCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTSOJUGI2TGOBUGQ). You are receiving this because you authored the thread.Message ID: ***@***.***>

[ecasey]

default timeout in apache2 and fpm was 30 seconds. increased to 90. seems to help. i'll conitnue monitoring and let you know.

…

-- Errol Casey ***@***.*** Read some Messages, Play Games, or Chat! Try my BBS - https://toolazy.synchro.net/bbs

On Tuesday, March 3rd, 2026 at 6:42 PM, Matthew Asham ***@***.***> wrote: awehttam left a comment [(awehttam/binkterm-php#156)](#156 (comment)) It sounds like a timeout to me, either PHP's maximum execution time or the proxy has a maximum time it waits for a response. If you're using PHP FPM there could also be another timeout occurring. The only thing I can do is change it so that the poll all command returns immediately, which means you would not see any response output from the poll itself. — Reply to this email directly, [view it on GitHub](#156 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AB5JJBUSLVQKWPMXWICFMIL4O5UXZAVCNFSM6AAAAACWA3NUMCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTSOJUGI2TGOBUGQ). You are receiving this because you authored the thread.Message ID: ***@***.***>