Problem with latest update on rolling channel

[justusstiernacken]

Hey @junkerderprovinz

i pulled the latest update on rolling channel. But the docker container wont start. The latest channel works normal.
Here is the log. Thanks for your help :)

[junkerderprovinz]

Thanks for the log, it showed what went wrong. The search index in /var/lib/opencloud/search/bleve-v5 was damaged: it still listed a segment file (000000000249.zap) that no longer existed, so OpenCloud could not load it. The search service then failed several times in a row, and when that happens OpenCloud stops the whole server. The :latest channel was not affected because OpenCloud 7 keeps its index in a different folder.

v1.3.1 checks the index before OpenCloud starts. If it cannot be opened, it is moved to bleve-v5.broken, OpenCloud starts with a fresh one, and all spaces are indexed again in the background. Update the container on the rolling channel and it should come up again. The log shows moved it to bleve-v5.broken and later search index rebuilt.

I have not found out yet how the file went missing. Killing OpenCloud in the middle of indexing a few hundred times never damaged the index, so a slow shutdown is unlikely to be the cause. Could you help narrow it down? Please keep the bleve-v5.broken folder for now and post:

1. the output of this command in the container console:

   stat -c '%y %s %n' /var/lib/opencloud/search/bleve-v5.broken/store/*
   

2. the file search/bleve-v5.broken/store/root.bolt from your data folder (it lists the index files, no file contents)
3. whether the data path is mapped through /mnt/user/... or /mnt/cache/..., and whether around the time it broke the server had a hard reset or power loss, an appdata backup or restore ran, or a second OpenCloud container used the same folder.

The crash itself is reported upstream as opencloud-eu/opencloud#3588.

I'm closing this since v1.3.1 fixes the crash. If it still doesn't start for you, or once you have the details above, just comment here and I'll reopen it.

[justusstiernacken]

Thanks for your reply. The Container not starting so i cant get any infos from console. I did a clean reinstall and it works now.

[justusstiernacken]

@junkerderprovinz
after reinstall to rolling 8.0.1 i tryed to copy all files over unraid filesystem. But after some time the "Personal" Point is missing. And i dont get it back. You have a tip?

[junkerderprovinz]

Hi @justusstiernacken, thanks for the update.

OpenCloud only takes over files copied straight into its Data folder if they land inside your account's folder, belong to nobody:users, and the container is stopped while you copy. If the folder of your personal space itself gets replaced or loses the hidden attributes OpenCloud keeps on it, OpenCloud stops listing it and Personal disappears. It does not repair that on its own.

Since you just reinstalled, a fresh start is the quickest way back:

1. If some of your files now only exist inside the OpenCloud data folder, copy them somewhere else first.
2. Docker tab, click the OpenCloud icon, Stop.
3. Delete or rename the config and data folders set in your template (by default /mnt/user/appdata/opencloud/config and /mnt/user/appdata/opencloud/data). This removes everything stored in OpenCloud, including accounts and shares.
4. Start the container and sign in once.
5. Copy your files in as described in the new README section "Personal" is gone after copying files into the Data folder. It walks through the terminal commands step by step. I ran them on Unraid 7.3.2 with OpenCloud 8.0.1.

Uploading through the web UI or the desktop client always works too, it's just slower for a big collection.

[justusstiernacken]

@junkerderprovinz

i tried both ways and they didnt work for me. I setup a fresh installation and tried with the desktop client. After a few GB the docker container is freezed and only a unraid reboot helped.

{"level":"error","service":"gateway","host.name":"fd645563282c","pkg":"rgrpc","traceid":"19cb647407e79c357def386e389d0a2b","user-agent":"grpc-go/1.83.2","from":"tcp://127.0.0.1:58814","uri":"/cs3.gateway.v1beta1.GatewayAPI/ListReceivedOCMShares","start":"28/Sep/2026:15:53:56 +0000","end":"28/Sep/2026:15:54:26 +0000","time_ns":29995191801,"code":"Canceled","time":"2026-09-28T15:54:26Z","line":"github.com/opencloud-eu/reva/v2@v2.50.0/internal/grpc/interceptors/log/log.go:69","message":"gateway: error calling ListReceivedOCMShares: rpc error: code = Canceled desc = context canceled"}
{"level":"error","service":"postprocessing","error":"context deadline exceeded","time":"2026-09-28T15:54:30Z","line":"github.com/opencloud-eu/opencloud/services/postprocessing/pkg/service/service.go:493","message":"failed to get consumer"}
{"level":"error","service":"search","error":"context deadline exceeded","time":"2026-09-28T15:54:31Z","line":"github.com/opencloud-eu/opencloud/services/search/pkg/service/event/service.go:244","message":"failed to get consumer"}
{"level":"error","service":"postprocessing","error":"context deadline exceeded","time":"2026-09-28T15:54:35Z","line":"github.com/opencloud-eu/opencloud/services/postprocessing/pkg/service/service.go:493","message":"failed to get consumer"}
{"level":"error","service":"search","error":"context deadline exceeded","time":"2026-09-28T15:54:36Z","line":"github.com/opencloud-eu/opencloud/services/search/pkg/service/event/service.go:244","message":"failed to get consumer"}
{"level":"error","service":"postprocessing","error":"context deadline exceeded","time":"2026-09-28T15:54:40Z","line":"github.com/opencloud-eu/opencloud/services/postprocessing/pkg/service/service.go:493","message":"failed to get consumer"}
{"level":"error","service":"search","error":"context deadline exceeded","time":"2026-09-28T15:54:41Z","line":"github.com/opencloud-eu/opencloud/services/search/pkg/service/event/service.go:244","message":"failed to get consumer"}
{"level":"error","service":"postprocessing","error":"context deadline exceeded","time":"2026-09-28T15:54:45Z","line":"github.com/opencloud-eu/opencloud/services/postprocessing/pkg/service/service.go:493","message":"failed to get consumer"}
{"level":"error","service":"search","error":"context deadline exceeded","time":"2026-09-28T15:54:46Z","line":"github.com/opencloud-eu/opencloud/services/search/pkg/service/event/service.go:244","message":"failed to get consumer"}
{"level":"error","service":"postprocessing","error":"context deadline exceeded","time":"2026-09-28T15:54:50Z","line":"github.com/opencloud-eu/opencloud/services/postprocessing/pkg/service/service.go:493","message":"failed to get consumer"}
{"level":"error","service":"search","error":"context deadline exceeded","time":"2026-09-28T15:54:51Z","line":"github.com/opencloud-eu/opencloud/services/search/pkg/service/event/service.go:244","message":"failed to get consumer"}
{"level":"error","service":"postprocessing","error":"Error publishing message to topic: nats: timeout","time":"2026-09-28T15:54:54Z","line":"github.com/opencloud-eu/opencloud/services/postprocessing/pkg/service/service.go:301","message":"unable to publish event"}
{"level":"fatal","service":"postprocessing","error":"fatal error: unable to publish event","time":"2026-09-28T15:54:54Z","line":"github.com/opencloud-eu/opencloud/services/postprocessing/pkg/service/service.go:136","message":"fatal error - exiting"}

[junkerderprovinz]

Hi @justusstiernacken, thanks for the log, and sorry this keeps getting in your way.

I tried to reproduce it on Unraid 7.3.2 with the same image (OpenCloud 8.0.1). I uploaded more than 10 GB the way the desktop client does it, many small files and several 1 GB files at once, with the Data folder on the cache and on the array. It did not freeze, and OpenCloud used less than 1 GB of RAM while doing it.

Your log shows that OpenCloud's internal message service stopped answering for about a minute, and after that OpenCloud shut itself down. That part is expected: Docker starts it again a few seconds later, and in my tests the uploads simply continued. So the crash itself is not what needed the reboot. When a container can't even be stopped anymore, the problem sits below OpenCloud, in the disk under the Data folder or in the server's memory.

1. Check where your Data folder is: Docker tab, click the OpenCloud icon, Edit, look at Data. It should start with /mnt/user/appdata/ or with the name of a pool on the Main tab, for example /mnt/cache/. Anywhere else, Unraid keeps it in RAM, and a few GB of uploads fill it until the server freezes.
2. On the Main tab, click the disk that holds that folder and check its SMART status.
3. If it freezes again and the web page still opens, go to Tools → Diagnostics and click Download before you reboot, and attach the zip here.

What exactly is in your Data field, how much RAM does your server have, and when it froze, was only OpenCloud stuck or the whole Unraid page too?

[justusstiernacken]

@junkerderprovinz

yeah its strange. It feels like as i start over and over again :D

The Server have 16GB Memory. Datashare ist mnt/user/opencloud and chache (ssd) is mnt/user/appdata.
I checked every disk of smart failures, but everything seems ok.

Today i tried again. I copied a Folder with many files to the opencloud folder on desktop to sync with client. After a few files opencloud freezed. i cant restart the opencloud container then and have to hardrestart the whole unraid server.

[junkerderprovinz]

Hi @justusstiernacken, thanks, that's the missing piece. /mnt/user/opencloud is a share on the array, and that's the setup that made big syncs stall on my own server too. OpenCloud keeps its message bus and file metadata in the Data folder and writes to them constantly, and during a large sync the array can't keep up. My last comment made it sound as if any /mnt/user/ path is fine, sorry about that.

Moving Data to the SSD fixed it for me:

1. Docker tab, click the OpenCloud icon, Stop.
2. If some files only exist in OpenCloud, copy them out of /mnt/user/opencloud first. Then rename that folder and your config folder (by default /mnt/user/appdata/opencloud/config).
3. Click the icon again, Edit, and set Data to a folder on the SSD, using the pool name from the Main tab, for example /mnt/cache/appdata/opencloud/data. Click Apply.
4. Sign in and let the desktop client sync again.

The background is in the README under Large-folder sync from the desktop client stalls or aborts. I'm closing this for now. If the SSD is too small for all your files, or it freezes again with Data on the SSD, just comment here with the size of the SSD and I'll reopen it.