If the document is not saved after modification and the browser is refreshed, the document will not be able to be loaded

User Support
1

Hi ,

If the document is not save after modification and the browser is refreshed, this document will never be able to be opened again.Why is this? Is there anything that needs to be done?

image
image1904Ă—727 10.2 KB

image
image1903Ă—756 8.93 KB

Best Regards.

2

Hi @Chenkai,

Thanks for posting.

If the document is not save after modification and the browser is refreshed, this document will never be able to be opened again.

So, you’re able to open/create a new document fine… but trying to reopen an old one causes COOL to get stuck on that screen?

Could you please give more info of your setup?

Q1] Can you give me your:

  • Help > About info?

Q2] How are you running Collabora? (OS? Browser version?)

Q3] Are there any errors appearing in the commandline when you try to open the old document? (File permission problems?)

3

Hi @Tex

repro steps:
1.Open a. docx document
2.Enter or modify some content, but close or refresh the browser without clicking save
3.If open this document again, Collaborateonline will not be able to display the document and will prompt error. The image above shows

Q1] Can you give me your:

Q2] How are you running Collabora? (OS? Browser version?)
OS： Linux version 5.15.0-58-generic (buildd@lcy02-amd64-101) (gcc (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #64-Ubuntu SMP Thu Jan 5 11:43:13 UTC 2023

Browser version：Microsoft Edge 113.0.1774.42

Q3] Are there any errors appearing in the commandline when you try to open the old document? (File permission problems?)

May 18 18:56:00 hkco coolwsd[463892]: kit-463892-439325 2023-05-18 18:56:00.850766 +0800 [ kitbroker_0cb ] TRC Purging dead sessions, have 1 active sessions.| kit/Kit.cpp:807
May 18 18:56:00 hkco coolwsd[463892]: kit-463892-439325 2023-05-18 18:56:00.850743 +0800 [ kitbroker_0cb ] TRC ToMaster-603: isDocLoaded state after loadDocument: false| kit/ChildSession.cpp:249
May 18 18:56:00 hkco coolwsd[463892]: kit-463892-439325 2023-05-18 18:56:00.850722 +0800 [ kitbroker_0cb ] ERR ToMaster-603: Failed to get LoKitDocument instance for [file:///tmp/user/docs/H5ZEK>
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850689 +0800 [ forkit ] TRC ppoll start, timeoutMicroS: 1000000 size 1| net/Socket.cpp:339
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850669 +0800 [ forkit ] TRC #13: setupPollFds getPollEvents: 0x1| net/Socket.hpp:876
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850584 +0800 [ forkit ] TRC cleanupChildren reaped 0 children to have 2 left: 463892 464703 | kit/ForKit.cpp:327
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850565 +0800 [ forkit ] TRC cleanupChildren with 2 children| kit/ForKit.cpp:292
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850549 +0800 [ forkit ] TRC forkLibreOfficeKit limit: 0| kit/ForKit.cpp:444
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850527 +0800 [ forkit ] TRC #13: Revents: 0x0| net/Socket.hpp:1316
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850509 +0800 [ forkit ] TRC #13: Starting handling poll events of forkit at index 0 (of 1): 0x0| net/Socket.cpp:435
May 18 18:56:00 hkco coolwsd[439325]: frk-439325-439325 2023-05-18 18:56:00.850474 +0800 [ forkit ] TRC Poll completed with 1 live polls max (1000000us)| net/Socket.cpp:357
May 18 18:56:00 hkco coolwsd[463892]: kit-463892-439325 2023-05-18 18:56:00.850434 +0800 [ kitbroker_0cb ] FTL Failed to load the document. Setting TerminationFlag| kit/Kit.cpp:1484
May 18 18:56:00 hkco coolwsd[463892]: Failed to load the document. Setting TerminationFlag
May 18 18:56:00 hkco coolwsd[463892]: kit-463892-439325 2023-05-18 18:56:00.850394 +0800 [ kitbroker_0cb ] TRC #16: Wrote 50 bytes of 50 buffered data| net/Socket.hpp:1457
May 18 18:56:00 hkco coolwsd[439295]: wsd-439295-464702 2023-05-18 18:56:00.850371 +0800 [ docbroker_0cb ] TRC ppoll start, timeoutMicroS: 5000000 size 2| net/Socket.cpp:339
May 18 18:56:00 hkco coolwsd[439295]: wsd-439295-464702 2023-05-18 18:56:00.850358 +0800 [ docbroker_0cb ] TRC #20: setupPollFds getPollEvents: 0x1| net/Socket.hpp:876
May 18 18:56:00 hkco coolwsd[439295]: wsd-439295-464702 2023-05-18 18:56:00.850343 +0800 [ docbroker_0cb ] TRC #35: setupPollFds getPollEvents: 0x1| net/Socket.hpp:876
May 18 18:56:00 hkco coolwsd[439295]: wsd-439295-464702 2023-05-18 18:56:00.850328 +0800 [ docbroker_0cb ] TRC Poll: current activity:

4

Thanks. For searchability, let me reproduce your info below:

Error

Failed to load the document. Please ensure the file type is supported and not corrupted, and try again.

About

COOLWSD version:
22.05.12.2 (git hash: c0a4f61)
LOKit version:
built on a great technology base
Collabora Office 22.05.11.1 (git hash: 20d7735)
Served by: Ubuntu 22.04.1 LTS 9aea8bba

Hmmm, I also notice you’re on:

  • 25.05.11

Mind updating to:

  • 25.05.14

Maybe that might have fixed your issue. :slight_smile:

Browser version：Microsoft Edge 113.0.1774.42

Edge? On Ubuntu? Hmmm, do other browsers cause this same thing too?

Failed to get LoKitDocument instance for [file:///tmp/user/docs/H5ZEK>

Hmmm… Doing a search for this error + file:///tmp/user/docs leads to some older Nextcloud/bugs (that were already fixed), like:

Hmmm… can you explain your current setup a little more?

So, you’re

  • running COOL on Ubuntu
  • Accessing COOL from a (different computer via Edge?)

Are you running COOL in a container or proxy or something else like that?

5

Hi @Tex

  • Are you running COOL in a container or proxy or something else like that?

    Yes,running COOL on Ubuntu.

  • Accessing COOL from a (different computer via Edge?)
    Yes,Google browser access is the same issue.

Our QA is using version 25.05.11
Our product uses 22.05.14.3, we tested, without this issue .Thank you for your reply.

Thanks

6

Our QA is using version 25.05.11
Our product uses 22.05.14.3, we tested, without this issue .Thank you for your reply.

So, if I’m reading this correctly, the issue was fixed in 22.05.14? :slight_smile:

7

Hi @Tex ,

Yes ,fixed in 22.05.14.3. But if change the server use, there is still a problem(Collabora is directly embedded in our own service, only one server is good, other servers still have problems)

But after I upgraded to 23.05.0.5, this problem appeared again

  1. Help > About info

    image
    image1135Ă—690 52.7 KB

  2. After the page is close or refresh, save is not clicked, and collaboraonline will still call the API of Save Document. At this time, the file saved is wrong. , I can’t open this file locally

Thanks.

8

Thanks. I’m pasting the About Info text from your image for future searchability:

COOLWSD version:
23.05.0.5 (git hash: b6cb979)

LOKit version:
built on a great technology base
Collabora Office 23.05.0.5 (git hash: 8b92c8b)
Served by: Ubuntu 22.04.2 LTS 1f3ed10d

But after I upgraded to 23.05.0.5, this problem appeared again

Hmmm. So that computer that was not working in:

  • 22.05.11

was working in:

  • 22.05.14.3

now re-broke in:

  • 23.05.0.5

And the ones that were broken the entire time, still weren’t fixed by updates? (Same OSes/symptoms/configs + everything else though?)

But if change the server use, there is still a problem(Collabora is directly embedded in our own service, only one server is good, other servers still have problems)

Hmmm… and no difference you can tell between the “good/working” server and the others?

Are the logs all showing those same errors you linked to in Post #2?

Random question:

  • Do you happen to have any odd characters/symbols in your filenames?
    • (I skimmed a look in your Post #3 image, and I see “.docx.docx”.)

This refresh problem is occurring in all documents?

9

Hi @Tex

  1. And the ones that were broken the entire time, still weren’t fixed by updates? (Same OSes/symptoms/configs + everything else though?)
    still weren’t fixed

  2. Hmmm… and no difference you can tell between the “good/working” server and the others?
    Good server and collaborative server are on the same network, which is the only difference

  3. Are the logs all showing those same errors you linked to in Post #2
    No error displayed

Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980790 2023-06-15 13:45:30.935277 +0800 [ admin ] TRC ppoll start, timeoutMicroS: 1000000 size 1| net/Socket.cpp:341
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980790 2023-06-15 13:45:30.935259 +0800 [ admin ] TRC #19: setupPollFds getPollEvents: 0x1| net/Socket.hpp:841
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980790 2023-06-15 13:45:30.935239 +0800 [ admin ] TRC Admin poll for 1000ms| wsd/Admin.cpp:546
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980790 2023-06-15 13:45:30.935216 +0800 [ admin ] TRC Message to admin console: cpu_stats 2| wsd/AdminModel.cpp:449
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980790 2023-06-15 13:45:30.935059 +0800 [ admin ] TRC #19: Handling poll events of admin at index 0 (of 1): 0x0| net/Socket.cpp:457
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980790 2023-06-15 13:45:30.935040 +0800 [ admin ] TRC #11: Handling events of wakeup pipe: 0x0| net/Socket.cpp:364
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980790 2023-06-15 13:45:30.934972 +0800 [ admin ] TRC Poll completed with 0 live polls max (750000us)(timedout)| net/Socket.cpp:360
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980773 2023-06-15 13:45:30.925248 +0800 [ prisoner_poll ] TRC ppoll start, timeoutMicroS: 1689300 size 2| net/Socket.cpp:341
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980773 2023-06-15 13:45:30.925231 +0800 [ prisoner_poll ] TRC #18: setupPollFds getPollEvents: 0x1| net/Socket.hpp:841
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980773 2023-06-15 13:45:30.925215 +0800 [ prisoner_poll ] TRC #17: setupPollFds getPollEvents: 0x1| net/Socket.hpp:841
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980773 2023-06-15 13:45:30.925197 +0800 [ prisoner_poll ] TRC #17: Handling poll events of prisoner_poll at index 0 (of 2): 0x0| net/Socket.cpp:457
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980773 2023-06-15 13:45:30.925176 +0800 [ prisoner_poll ] TRC #18: Handling poll events of prisoner_poll at index 1 (of 2): 0x0| net/Socket.cpp:457
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980773 2023-06-15 13:45:30.925155 +0800 [ prisoner_poll ] TRC #7: Handling events of wakeup pipe: 0x0| net/Socket.cpp:364
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980773 2023-06-15 13:45:30.925073 +0800 [ prisoner_poll ] TRC Poll completed with 0 live polls max (5000000us)(timedout)| net/Socket.cpp:360
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895894 +0800 [ forkit ] TRC ppoll start, timeoutMicroS: 1000000 size 1| net/Socket.cpp:341
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895877 +0800 [ forkit ] TRC #13: setupPollFds getPollEvents: 0x1| net/Socket.hpp:841
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895854 +0800 [ forkit ] TRC cleanupChildren reaped 0 children to have 2 left: 1028767 1030207 | kit/ForKit.cpp:326
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895830 +0800 [ forkit ] TRC cleanupChildren with 2 children| kit/ForKit.cpp:291
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895812 +0800 [ forkit ] TRC forkLibreOfficeKit limit: 0| kit/ForKit.cpp:442
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895787 +0800 [ forkit ] TRC #13: Handling poll events of forkit at index 0 (of 1): 0x0| net/Socket.cpp:457
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895766 +0800 [ forkit ] TRC #11: Handling events of wakeup pipe: 0x0| net/Socket.cpp:364
Jun 15 13:45:30 hkco coolwsd[980774]: frk-980774-980774 2023-06-15 13:45:30.895675 +0800 [ forkit ] TRC Poll completed with 0 live polls max (1000000us)(timedout)| net/Socket.cpp:360
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980789 2023-06-15 13:45:30.458440 +0800 [ websrv_poll ] TRC ppoll start, timeoutMicroS: 5000000 size 0| net/Socket.cpp:341
Jun 15 13:45:30 hkco coolwsd[980766]: wsd-980766-980789 2023-06-15 13:45:30.458418 +0800 [ websrv_poll ] TRC #5: Handling events of wakeup pipe: 0x0| net/Socket.cpp:364

  1. This refresh problem is occurring in all documents?
    Yes

  2. Do you happen to have any odd characters/symbols in your filenames?(I skimmed a look in your Post #3 image, and I see “.docx.docx”.)
    No OD characters/symbols，But we do have files named “. docx. docx”

  3. After the page is close or refresh, save is not clicked, and collaboraonline will still call the API of Save Document. At this time, the file saved is wrong. , I can’t open this file locally
    Is this the problem?

1 Like
10

After the page is close or refresh, save is not clicked, and collaboraonline will still call the API of Save Document. At this time, the file saved is wrong. , I can’t open this file locally

Is this the problem?

Hmmm. In last week’s Community Meeting, I brought this topic up.

Some ideas were spitballed.

Like it could be the document gets stuck in some sort of limbo “partial lock” or “corrupt save state”.

COOL might still believe the file is being modified/saved, so won’t let a new user open until that safely completes.

@andras.timar , think you could help debug this one? :slight_smile: