Switching storage location to S3....not working at all. Again.

[MaNiAc LRSC]

We are in the process of switching storage endpoints from local storage to S3 endpoints.

As a first test, we encountered an error when switching the "User emojis" storage location. There was a message stating that another migration is still in progress, which was not the case as far as we can see. As a result, none of the "user emoji" files were migrated to the new storage location.
After this error, new posts tried to use emojis at the new storage location (which did not have any emojis uploaded/migrated) while pre-switching posts still referenced to the prior emoji storage location.

At the end of the day we had following problems:

1. Files from old to new storage location were not migrated
2. Old posts were not modified to use the new storage location

To fix this problem short-term, we copied the "user emoji" files manually to the S3 storage as well to ensure the users don't see broken links.

 

Questions:

1. We already tried such a migration a few months before and hat severe problems as well. Support ticket at this time was not solved properly, only non-helping hints were given.
   Now we tried again but not even the "user emojis" were moved properly...what will happen when we try to migrate the 900GB of attachments?! 
   What needs to be done to migrate storage locations properly?
2. Is there a way to start a migration by command line to prevent e.g. timeouts?
3. What needs to be done so that old posts /old data uses the new storage location as well?

 

Edited February 17, 2022 by MaNiAc LRSC

[Jim M]

When you attempted the migration did you select for the system to migrate the files for you or was that unselected? If so, did the background tasks ever run or become present? 

Currently, there are no background tasks present which makes me think that the setting to have the system move things was disable and thus no images were moved. As if there was an error, these tasks would still be present.

You also will want to take a look at your System Log as there are over 7GB worth of Elastic Search connection errors there. This is making it hard to find any previous errors as well.

[Skillshot]

We did a migration for calendar events at first which went ok'ish (with some minor problems). The migration has completed, there where no tasks pending. When we switched the storage location for custom emojis including the "move data" option, there was an error displayed stating that another migration is still in place and therefore it is not possible to switch locations. Despite that, the location was switched for new posts leading to broken images because the images had not been transfered to the new location.

Is there a way to manually re-trigger migrations in such a case?

Is there some kind of cli tooling available to run such tasks manually without involving the frontend?

Is there an easy way other than poking around directly in the database to just generate a list of file (and locations) that a specific movement would touch?

[Marc Stridgen]

Hello,

 

Quote

Is there a way to manually re-trigger migrations in such a case?

There would be no way in which to retrigger those. You would likely be best to revert to the backup taken before this process. If you are then still having issues, we can create a ticket to take a look for you if needed.

Quote

Is there some kind of cli tooling available to run such tasks manually without involving the frontend?

There is no such tool available. It is done with the background tasks

Quote

Is there an easy way other than poking around directly in the database to just generate a list of file (and locations) that a specific movement would touch?

There isnt really, unfortuanetly.