Odrive not working, Finder extension stopped working on Catalina 10.15.5

Troubleshooting
1

Hi

After the initial installation 3 days ago on an iMac (late 2017, 24GB RAM, 2TB ) and sync of my Google Drive with roughly 370GB all seemed fine. So I also migrated a second iMac to odrive.

For an unknown reason syncing stopped on Mac #1. I have obvious gaps between Google Drive and iMac #1. CPU usage is at 100% even after restarting, Finder extension stopped working.

Please advice

Thanks
Ray

2

Hi @ray.popoola,
Sorry to hear you are having some issues.

Is the odrive tray menu working? If so, can you send a diagnostic from it so that I can take closer look?

3

Hi @Tony

Thanks for following up.
Sometimes the tray menu does not come up or when being lucky.
Here the diagnostic filecurrent_odrive_status.txt (2.0 KB)

Best
Ray

4

Hi @ray.popoola,
Thanks for the follow-up. I took a look at the diagnostic and there is a lot of activity happening, which can spike the CPU.

Can you walk me through the actions you are currently taking with odrive? Are you attempting to download everything in your Google Drive to your local system?

If your current goal is bulk download of your remote data, we can try to scope the amount of work odrive is doing to just that, until it is done. This will reduce the overhead and should allow for smoother operation. To do this, perform the following steps:

  • Exit odrive
  • Double-click to open the “odrive_user_general_conf.txt” in the root of your odrive folder
  • Change "disableFSEvents": false, to "disableFSEvents": true, and then save the file. This will disable odrive’s monitoring of the local filesystem events, which can get pretty noisy during bulk download operations.
  • Start odrive again
  • Once the tray menu is up, go to the top item in the menu (it will say “Ready to sync new changes” or “Syncing changes”) and select “Disable background scanning”.
  • Proceed with your download operation.

Let me know if you notice any changes in behavior after doing this.

5

Thanks @Tony

Basically I already downloaded most of the files.
I followed your instructions and restarted and now it seems to work. Under what circumstances should I reverse above changes?

Best
Ray

6

Hi @ray.popoola,
You should be able to reverse them after you are done with the bulk download. We are also working on an adjustment that will reduce some noise in the events, so disabling may not be needed next time.

7

Thanks, @Tony

One problem remains. After unlocking this morning the tray wasn’t accessible, CPU on about 100% and 5 odrive Finder extensions running. Syncing didn’t work.

So I killed those processes and restarted the app. After that odrive was back to normal.

I’m not sure where we are at. I’m testing the trial for 7 days and fixing issues for 4 days…

Please advice

Thanks

Ray

8

@Tony

On Machine #2 different: odrive tray is accessible, but odrive is not syncing. Here is the status report: current_odrive_status.txt (8.8 KB)

I killed all odrive processes and restarted. Then it synced.
What know issues are there I need to know about?

Thanks
Ray

9

Hi @ray.popoola,

Thanks for the information! Everyone’s data is different and you look to have a lot of what we call “sprawling” folder structures. These are typical in photo and video editing type of software, where they tend to have these ridiculously complex data structures with thousands and thousands of folders. You’ll see these, for example, in software like Adobe Lightroom, Apple Photos, and Scrivener.

The most difficult use cases to make really smooth are those dealing with bulk operations, like bulk downloading, especially against sprawling structures, as mentioned above.

One of the big ideas behind odrive’s sync engine is the idea of “progressive” sync, where you can get to and sync the specific data you need at that time without having to sync down the entirety of your remote data. The ability to do this can cut out a ton of unnecessary overhead because it allows odrive (and users) to focus on a specific subset of the data instead of having to monitor and track everything. Because this is sync and not basic import/export, during and after bulk operations odrive still needs to constantly look at everything, on both the local and remote side, to make sure it is all in sync. This is a constant, never-ending task that odrive has to perform. The larger the scope of the data that odrive needs to monitor, the more overhead you are likely to have.

This all being said, we still want to be able to make these difficult scenarios smooth. The information you provided yesterday actually pointed out a situation where we can potentially get into a feedback loop when bulk downloading thousands of these sprawling types of folders in rapid succession. Here is a test version that addresses that situation. I am hoping you can run to see if it solves a lot of the CPU heat you are seeing: https://www.odrive.com/s/aea2a683-cacc-4d4c-b5e9-44ea967fd55b-5ef2179d

  • So that I can further understand your method of odrive use and use cases better, can you walk me through what actions you are taking prior to recognizing that things are not syncing?

  • Can you also describe the behavior when you notice that syncing is not working? Is this for downloads, uploads, or both?

The status files you sent over actually indicate that things are syncing, but it is a bit deep in the bowels of some of these folder structures, so it may not be readily apparent that something is actually happening. For example, in the last odrive status file you sent, it looks like it was working its way through one of these Lightroom structures (you can see it under “Active (4):” in the status file).

There are some tricks we can employ, as well, to cut down on unnecessary processing, like identifying caches or what I call “throw-away” files and folders that could be excluded from sync using our custom blacklist feature. It’s just a matter of identifying the right stuff.

10

Hi @Tony

Thanks for your insights.
I identified a Lightroom file that never seemed to stop syncing and blacklisted that file. I hope that solves the issues.

The way I synced with a previous product was an almost complete sync, rest being kind of an archive in the cloud. I will reconsider this with odrive.

There were no extraordinary steps prior realizing that files didn’t sync. It were files I use on a regular base and they were not there. Issues where predominately with files not being downloaded.

11

Thanks for the follow-up @ray.popoola.

Have you setup “folder sync rules” for the folders that you are actively working with? These rules will allow you to specify that you want any new content auto-downloaded. This can be important for the types of structures you are working with, since they tend to be fairly dynamic in terms of new folders added within the structures.

12

@Tony, thanks.
I’ve set a general rule for Google Drive in the beginning to download all files. Now I’m reversing that for folders that serve more as an archive and will exclude them from syncing.

1 Like
13

Hi @ray.popoola,
I just wanted to let you know that we have an official release of the test version I posted here earlier (which includes the optimization to reduce the overhead I was seeing during your bulk download, along with other improvements).