31 Aug 2021 |
| [anoa] joined the room. | 07:33:33 |
| ScorchedMuffin joined the room. | 18:57:27 |
1 Sep 2021 |
| philfree joined the room. | 17:32:11 |
| oss_guy joined the room. | 23:01:41 |
2 Sep 2021 |
niquola | Hi guys, any plans to add Wiki-like links between notes - [[..autocomplete notes or plugin system so we can contribute it? | 14:04:19 |
| Jason van Zyl joined the room. | 14:17:24 |
d33tah | hi! quick question: what are the disadvantages of sqlite as the database backend? hedgedoc recently started crashing and the logs mention "error: operation revision not in history" | 14:29:27 |
David [🦔-backend] | In reply to @niquola:matrix.org Hi guys, any plans to add Wiki-like links between notes - [[..autocomplete notes or plugin system so we can contribute it? We plan a few features in this direction for HedgeDoc 2, see https://github.com/hedgedoc/hedgedoc/issues?q=is%3Aissue+is%3Aopen+sort%3Aupdated-desc+label%3A%22feature%3Anote+organization%22
The closest thing is probably https://github.com/hedgedoc/hedgedoc/issues/1001
| 14:29:47 |
David [🦔-backend] | In reply to @d33tah-test:matrix.org hi! quick question: what are the disadvantages of sqlite as the database backend? hedgedoc recently started crashing and the logs mention "error: operation revision not in history" SQLite does not scale as well as PostgreSQL or MariaDB and has no high availability. It should not cause HedgeDoc to crash, so I'd be interested in more details. Could you post a snippet of the log that shows the crashß | 14:32:27 |
David [🦔-backend] | In reply to @d33tah-test:matrix.org hi! quick question: what are the disadvantages of sqlite as the database backend? hedgedoc recently started crashing and the logs mention "error: operation revision not in history" * SQLite does not scale as well as PostgreSQL or MariaDB and has no high availability. It should not cause HedgeDoc to crash, so I'd be interested in more details. Could you post a snippet of the log that shows the crash? | 14:32:33 |
David [🦔-backend] | * SQLite does not scale as well as PostgreSQL or MariaDB and has no networking features or high availability. It should not cause HedgeDoc to crash, so I'd be interested in more details. Could you post a snippet of the log that shows the crash? | 14:33:03 |
d33tah | In reply to @davebloggt:kif.rocks SQLite does not scale as well as PostgreSQL or MariaDB and has no networking features or high availability. It should not cause HedgeDoc to crash, so I'd be interested in more details. Could you post a snippet of the log that shows the crash? is there a standard log location for your Docker images? the stdout mostly mentions user deserialization | 14:34:00 |
David [🦔-backend] | We log to stdout, so the docker logs should contain everything | 14:34:37 |
d33tah | hm, stdout doesn't seem to have details | 14:38:59 |
d33tah | I looked into queries and found that this is the kind of query that fails:
curl 'https://scrubbed/socket.io/?noteId=scrubbed&EIO=3&transport=polling&t=NkcyECC&sid=scrubbed' -H 'User-Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:91.0) Gecko/20100101 Firefox/91.0' -H 'Accept: */*' -H 'Accept-Language: en-US,en;q=0.5' --compressed -H 'Referer: https://scrubbed' -H 'Content-type: text/plain;charset=UTF-8' -H 'Origin: https://scrubbed' -H 'Authorization: Basic scrubbed' -H 'Connection: keep-alive' -H 'Cookie: connect.sid=scrubbed; loginstate=true; indent_type=space; space_units=4; keymap=sublime; userid=scrubbed' -H 'Sec-Fetch-Dest: empty' -H 'Sec-Fetch-Mode: cors' -H 'Sec-Fetch-Site: same-origin' -H 'Sec-GPC: 1' -H 'Pragma: no-cache' -H 'Cache-Control: no-cache' --data-raw '1:2'
| 14:39:17 |
d33tah | getting error 400 | 14:39:36 |
d33tah | "Session ID unknown" | 14:39:52 |
David [🦔-backend] | This can happen when the session has expired | 14:41:25 |
David [🦔-backend] | Did your users report any problems? | 14:41:32 |
d33tah | I'm the main user and the admin for this instance | 14:41:43 |
d33tah | it only happens when I delete sections | 14:41:50 |
d33tah | it doesn't make sense for session expire, surrounding queries work fine | 14:42:14 |
d33tah | but this 400 makes hedgedoc mark itself as offline, then i need to wait a few seconds and my latest change is undone | 14:42:39 |
David [🦔-backend] | So you can edit a note normally, then you delete a section of text and the editor in your browser is offline for a few seconds? | 14:43:49 |
Erik [🦔-dev] | In reply to @d33tah-test:matrix.org
I looked into queries and found that this is the kind of query that fails:
curl 'https://scrubbed/socket.io/?noteId=scrubbed&EIO=3&transport=polling&t=NkcyECC&sid=scrubbed' -H 'User-Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:91.0) Gecko/20100101 Firefox/91.0' -H 'Accept: */*' -H 'Accept-Language: en-US,en;q=0.5' --compressed -H 'Referer: https://scrubbed' -H 'Content-type: text/plain;charset=UTF-8' -H 'Origin: https://scrubbed' -H 'Authorization: Basic scrubbed' -H 'Connection: keep-alive' -H 'Cookie: connect.sid=scrubbed; loginstate=true; indent_type=space; space_units=4; keymap=sublime; userid=scrubbed' -H 'Sec-Fetch-Dest: empty' -H 'Sec-Fetch-Mode: cors' -H 'Sec-Fetch-Site: same-origin' -H 'Sec-GPC: 1' -H 'Pragma: no-cache' -H 'Cache-Control: no-cache' --data-raw '1:2'
That query instructs socket.io to use long-polling which is inferior to websockets as transport method. Maybe there's something faulty with your reverse proxy configuration regarding websockets? | 14:44:00 |
hanser | I also had issues like this with my reverse proxy. Could you check it without the proxy? | 14:44:41 |
David [🦔-backend] | In reply to @erik:michelson.eu That query instructs socket.io to use long-polling which is inferior to websockets as transport method. Maybe there's something faulty with your reverse proxy configuration regarding websockets? socketio seems to do these queries sometimes, even if the websocket works | 14:44:51 |
d33tah | hm, there seems to be no websocket connection made according to the browser | 14:45:42 |
Erik [🦔-dev] | Yes, but it still could be a thing. | 14:45:42 |
d33tah | is there a recommended configuration for nginx that should work? | 14:45:52 |