Do you want to: We previously updated our OnlyOffice for Liferay in order to solve a problem of documents not saving to a version history. Upon updating new versions started saving whenever a change was made however none of the changes are actually being saved to the new version of documents. These documents are saving to new versions with their own version numbers however they do not retain any changes made. The changes are still visible when we go to OnlyOffice Edit but not when we download the document or go to a specific version that should have had the changes.
Document Server version: v7.1.1
OS: CentOS
Browser version: Chrome
Liferay Community Edition Portal 7.3.5 CE GA6
Connector App v2.2
Hello @BShaver
Could you please provide us with whole Document server logs folder? it’s located here:
docker: /app/onlyoffice/DocumentServer/logs/documentserver/
deb\rpm: /var/log/onlyoffice/documentserver/
Also, if it’s possible, please reproduce the situation and make a videofile.
I’ve attached a video showing that I can enter onlyoffice edit, make changes, and then a screenshot after showing that the version history is not updated, neither is the preview. Additionally I cannot upload the textfile for the out logs as it is not a compatible file time. However I do see repeated errors in the logs.
Hello @BShaver
According to provided logs the issue is related to your SSL certificates. Probably intermediate certificate is missed in the chain. Please double check that both sides (Document server and Liferay) don’t have any issues with certificates. If servers have free access to the Internet, you can check SSL here:SSL Checker
I replaced both Liferay and Document server certificates with the recently renewed certificates that we have and made sure that the intermediate certificate was included as well with both .crt files and the .pfx file that liferay uses. After confirming both of these I restarted the related services and was still experiencing the same issue of changes not being saved to version even when a document is checked in with a minor or major version change selected. I am grabbing the logs from this time as well.
The documentserver logs are repeating the same log every minute:
“[ERROR] nodeJS - [AMQP] Error: connect ECONNREFUSED 127.0.0.1:5672
at TCPConnectWrap.afterConnect [as oncomplete] (net.js:1107:14)”
Additionally the nginx error log is alternating between the following entries:
*102 no live upstreams while connecting to upstream
*102 connect() failed (111: Connection refused) while connecting to upstream, client:
Is it possible to provide us with whole Document server logs folder? The provided log contains error entry which is related to RabbitMQ: Ports which must be opened for ONLYOFFICE Docs - ONLYOFFICE
Please double check that this port isn’t occupied by other app.
One more thing. Please check this folder on the Document server side: /var/lib/onlyoffice/documentserver/App_Data/cache/files/forgotten/
Are there any files?
Currently port 5672 is being used by amqp, which I believe is used by RabbitMQ, and currently there are roughly 100+ files in the forgotten folder. As for the log folder I am unable to upload the zip as I get this message when I try to. “Sorry, the file you are trying to upload is not authorised (authorised extensions: jpg, jpeg, png, gif, heic, heif, webp).”
Hello @BShaver
Sorry for the late reply. Please place logs to any external storage and provide us with download link.
The below link should contain the documentserver logs. Some of the information such as the IPs and web addresses have been replaced while retaining the same formatting.
https://drive.google.com/file/d/1_zJNzRDKjX-5NsGZykD7znqoK9rrYOmc/view?usp=sharing
It’s a little bit strange. The latest log entry is dated as January, 17th. Could you please check out out.log files in docservice and converter folders? are there more recent error entries?
That was the date which I pulled the log entries from the server, which should have also been after we were testing the issue, confirming certificates, as well as restarting the server. However, I can pull more recent logs and update the google drive link if it would be better to have.
Yes, please. We need the recent logs to check out the issue.
I’ve uploaded recent logs pulled from today.
https://drive.google.com/file/d/1AHuKt3B-F5Hn_SWw2wRVv554_GvNV0Qo/view?usp=sharing
I’m a little bit confused. I don’t see timestamps in error.log file, but I see a new error entry. Is it possible to contact me via PM and provide not edited logs (entire folder, all logs)?
Meantime, converter out.log file shows error entries with RabbitMQ. Please go to the host and run service rabbitmq-server status > please show us the result.
Also please run ls -la /etc/ > check output result for rabbitmq folder, please let us know the result.
One more test, please run ls -la /etc/rabbitmq ands show us the result.
In order the results are:
drwxr-sr-x 2 root rabbitmq 6 Mar 2 2021 rabbitmq
While looking into the rabbitmq service I also found that it was repeated failing and looping.
The full message from that log is as follows:
{Kernel pid terminated,application_controller,{application_start_failure,credentials_obfuscation,{{shutdown,{failed_to_start_child,credentials_obfuscaton_svc,{undef,[{crypto,macN,[hmac,sha256,<<206,5,237,104,252,49,205,197,35,230,167,152,214,192,183,27,225,208,164,3,220,180,206,50,57,86,138,56,100,62,170,6,174,179,40,88,105,56,234,23,211,73,175,108,240,183,149,97,30,238,209,190,108,155,221,228,220,141,198,1,111,24,28,139,232,126,162,235,30,196,34,130,247,213,146,190,104,185,82,185,10,216,132,130,239,60,6,127,189,237,185,105,197,254,23,124,192,99,21,197,244,2,64,59,218,211,146,147,133,160,241,152,38,165,138,63,87,134,46,58,176,36,186,119,62,102,185,152>>,[<<178,91,33,56,11,18,233,65,157,24,46,182,251,99,203,221>>,<<0,0,0,1>>],32],[]},{credentials_obfuscation_pbe,xor_sum,7,[{file,\src/credentials_obfuscation_pbe.erl},{line,160}]},{credentials_obfuscation_pbe,blocks,10,[{file,\src/credentials_obfuscation_pbe.erl},{line,152}]},{credentials_obfuscation_pbe,make_key,5,[{file,\src/credentials_obfuscation_pbe.erl},{line,94}]},{credentials_obfuscation_pbe,encrypt,5,[{file,\src/credentials_obfuscation_pbe.erl},{line,76}]},{credentials_obfuscation_svc,check,3,[{file,\src/credentials_obfuscation_svc.erl},{line,163}]},{credentials_obfuscation_svc,init_state,0,[{file,\src/credentials_obfuscation_svc.erl},{line,133}]},{gen_server,init_it,2,[{file,\gen_server.erl},{line,374}]}]}}},{credentials_obfuscation_app,start,[normal,[]]}}}}
Troubleshooting this message led to multiple posts stating that erlang needed to be updated except the erlang in the server is both updated and running in this instance.
It’s just crossed my mind that you are still using outdated version of Document server - v.7.1.1.
Would you mind preparing whole server backup\snaphot and update Document server to v.7.2.2?
This way we will check out if the issue persists on the latest version.
We had previously done the update and the issue persists, as well as causing additional issues with the liferay onlyoffice connector portlet which is why we had rolled back to a stable version we know was previously working correctly.
Did you perform the update to v.7.2.2 previously? As far as I understand, you have updated to v.7.1.1. Is it possible run the update process one more time (to current version of Document server)? Also please pay attention that we released connector app v.2.3.0. It’s better to trouble shoot the situation on the latest component versions.
Apologies for the late response, in order to better understand how everything was working I rebuilt the setup of Liferay version 7.3.5 CE GA6, the v2.3.0 connector app, and the v7.3.0 onlyoffice documentserver, following the instructions here. Currently when I go to edit with only office I get a blank page
Followed by an error message stating that the connection was interrupted. I’ve checked to make sure nginx, rabbitmq, postgresql are all running. Attached here is the google drive link to the document server logs in a zip. Hopefully if I can get this instance resolved then I can go over the production server and fix/rebuild what is needed.
Additionally when inspecting the web page via f12, the console loops with this error message.
