Reminder crashes

[tschlue]

After upgrading yesterday Mattermost to 5.14 and the reminder to 0.4.4 today the plugin was stopped (by itself of or mattermost) due to a lot of errors.

{"level":"error","ts":1566374506.8136456,"caller":"api4/websocket.go:28","msg":"websocket connect err: websocket: request origin not allowed by Upgrader.CheckOrigin"}
{"level":"error","ts":1566374506.8137093,"caller":"mlog/log.go:173","msg":"Failed to upgrade websocket connection","path":"/mattermost/api/v4/websocket","request_id":"xphyixw7htf47e7p88etupkdfh","ip_addr":"10.144.156.248","user_id":"6bwcgiotw3fk9dr3mzkozwxa5o","method":"GET","err_where":"connect","http_code":500,"err_details":""}
{"level":"error","ts":1566374509.8402839,"caller":"api4/websocket.go:28","msg":"websocket connect err: websocket: request origin not allowed by Upgrader.CheckOrigin"}
{"level":"error","ts":1566374509.8403947,"caller":"mlog/log.go:173","msg":"Failed to upgrade websocket connection","path":"/mattermost/api/v4/websocket","request_id":"kd4m737sfjre3p51mzyukmgpzc","ip_addr":"10.144.156.248","user_id":"6bwcgiotw3fk9dr3mzkozwxa5o","method":"GET","err_where":"connect","http_code":500,"err_details":""}
{"level":"error","ts":1566374537.4792485,"caller":"mlog/log.go:173","msg":"Plugin failed to ServeHTTP, RPC call failed","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind","error":"unexpected EOF"}
{"level":"error","ts":1566374537.4807625,"caller":"mlog/log.go:173","msg":"Action integration error","path":"/mattermost/api/v4/posts/ghyqpraj9fnn5qxfski4qg65dy/actions/zr6qd3f9kfyfucys5yddtw15cc","request_id":"ccetx9ukppbpfnaimszc5wdnbc","ip_addr":"10.144.157.66","user_id":"7rehqjww43r4tyo5mrmewj7tfc","method":"POST","err_where":"DoActionRequest","http_code":400,"err_details":"status=500"}
{"level":"error","ts":1566374550.488675,"caller":"mlog/log.go:173","msg":"Plugin failed to ServeHTTP, RPC call failed","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind","error":"connection is shut down"}
{"level":"error","ts":1566374550.4893534,"caller":"mlog/log.go:173","msg":"Action integration error","path":"/mattermost/api/v4/posts/ghyqpraj9fnn5qxfski4qg65dy/actions/i49sh7yej7n3dci9db3kkbm65y","request_id":"1c6943x8eb8yxmbjbfwquqd1ue","ip_addr":"10.144.157.66","user_id":"7rehqjww43r4tyo5mrmewj7tfc","method":"POST","err_where":"DoActionRequest","http_code":400,"err_details":"status=500"}
{"level":"error","ts":1566374555.4889648,"caller":"mlog/log.go:173","msg":"Plugin failed to ServeHTTP, muxBroker couldn't Accept request body connection","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind","error":"timeout waiting for accept"}
{"level":"error","ts":1566374555.4889503,"caller":"mlog/log.go:173","msg":"Plugin failed to ServeHTTP, muxBroker couldn't accept connection","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind","serve_http_stream_id":4,"error":"timeout waiting for accept"}
{"level":"error","ts":1566374561.0113795,"caller":"plugin/health_check.go:89","msg":"Health check failed for plugin com.github.scottleedavis.mattermost-plugin-remind, error: Plugin RPC connection is not responding"}
{"level":"error","ts":1566374561.011434,"caller":"mlog/log.go:173","msg":"RPC call OnDeactivate to plugin failed.","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind","error":"connection is shut down"}
{"level":"warn","ts":1566374561.0114772,"caller":"plugin/hclog_adapter.go:51","msg":"error closing client during Kill","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind","wrapped_extras":"errconnection is shut down"}
{"level":"warn","ts":1566374561.0114977,"caller":"plugin/hclog_adapter.go:53","msg":"plugin failed to exit gracefully","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind"}

[tschlue]

I realised that error stops appearing. A user has used the ip address of the mattermost server instead of the url. Other plugins didn't have a problem with it.

[tschlue]

Again

Looks like it wasn't the problem.

[scottleedavis]

@tsclue what is the type of server? Docker, manual install,? This looks like a big error and I'd like to reproduce it.

[tschlue]

Manual

[tschlue]

I think it is still the problem when connection URL not fits correctly. I moved the mattermost url from https://server.com to https://server.com/mattermost but the old url is rewritten to the new one.

[scottleedavis]

@tschlue I setup a docker and debian install of 5.14 this evening, and although had some issue with the buttons not working with debian (see issue #148 ), the plugin installed and ran without error on both. I am curious about what might be going on with the url rewrite you mentioned, though I don't think that would cause the plugin crash.

What operating system is this running on?
Does matterpoll work?

@lieut-data (pardon the ping), is this a trivial or known issue to fix? I looked here, however I didn't find anything specifically addressing plugin crashes. e.g. Plugin failed to ServeHTTP, RPC call failed, Plugin failed to ServeHTTP, muxBroker couldn't accept connection

[tschlue]

@scottleedavis Matterpoll works fine. Mattermost is running on centos 7.

Today only one error occurred
{"level":"error","ts":1566539999.695455,"caller":"mlog/sugar.go:23","msg":"No target found for reminder","plugin_id":"com.github.scottleedavis.mattermost-plugin-remind"}
Looks like it's connect with the url the user using in the client.
Did you look for that?

[scottleedavis]

@tschlue that error is different than the original one you posted. It should only occur once the plugin is running though. is the plugin running currently with just that error on triggered reminder? Could you share the /remind ..... string you used?

[tschlue]

@scottleedavis at the moment, yes. I wrote all user to correct the url path. So it could be still the reason for the errors. Which /remind ... string do you mean? Maybe I don't know beacause it's used by an user.

[scottleedavis]

@tschlue I wanted to check what the reminder being used was, e.g. /remind me to drink water at 3pm . Do you have something that was used to create the error?

[tschlue]

@scottleedavis No, I don't. But it could come from another user. I can't reproduce these errors :(.

[scottleedavis]

OK, since the crashing is over and no reproduction of the errors are currently possible do you feel we can close this issue? We can open another if anything comes up.

[tschlue]

I think this is okay.