Replies: 3 comments 10 replies
-
|
Immich automatically de-duplicates your (identical) assets when uploading assets through the api (so either immich-cli, immich-go, mobile, ...). So this shouldn't be a problem at all for you actually, you won't end up with duplicates in immich. |
Beta Was this translation helpful? Give feedback.
-
|
Yes I know. But it's a massive amount of extra pictures. They might be from relatives or they might be from me. Going through them will takes ages so I try to use an automatic way first by using this approach. If I upload them and realise they are unwanted then I would have screwed my current setup. |
Beta Was this translation helpful? Give feedback.
-
|
I am not sure I follow. In your original post you're saying
which is incorrect. Immich deduplicates based on identical file content, not metadata. |
Beta Was this translation helpful? Give feedback.
-
|
On the server, yes. At the upload level, the device id is involved.
Uploading the same picture exactly from 2 devices would result in 2 uploads. Yes these would be deduplicated later on the server but not at upload level
What I’m trying to simulate here is an upload level de duplication by overriding the deviceId. But we are getting quite away from the original questions that is whether it is possible to get a list of deviceId or not. I’ve seen few suggestions around but none seemed to work.
…On 7 Sep 2024 at 11:26 +0200, Daniel Dietzler ***@***.***>, wrote:
I am not sure I follow. In your original post you're saying
> My understanding is that deduplication is also based on the device Id
which is incorrect. Immich deduplicates based on identical file content, not metadata.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
Got you. I assumed the asset was uploaded then rejected there. What you're saying is that there is 2 different process
By curiosity why are the checksum/id used in 1 and 2 not simply the same ? Is it because computing the one used server side is more costly and that approach improves performance ? In my case, it means the original question becomes irrelevant. Thank you ! |
Beta Was this translation helpful? Give feedback.
-
Because we haven't gotten around to implementing that yet ;) |
Beta Was this translation helpful? Give feedback.
-
Yeah makes sense again. |
Beta Was this translation helpful? Give feedback.
-
Actually last question : Is there any documentation on what kind of hash is used server side ? Any way this could be computed by myself with some programming involved of course |
Beta Was this translation helpful? Give feedback.
-
|
We use an sha1 hash of the file contents. |
Beta Was this translation helpful? Give feedback.
-
|
Amazing. Thanks again
…On 7 Sep 2024 at 15:23 +0200, bo0tzz ***@***.***>, wrote:
We use an sha1 hash of the file contents.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
Hello,
Long story short, i'm trying to re-upload a mess of old pictures that I have. I suspect most of them are already in Immich, but must have been uploaded from a different device.
When I run immich-cli, or immich-go, most of them would be reuploaded (like 99% of them). It might be correct and expected, but I doubt it.
My understanding is that deduplication is also based on the device Id, and Immich-go offers the possibility to override it. So is it possible to get a list of all device Ids that were used to upload pictures in the past in Immich ? If I test with each of them and still get the same upload rate, then I would be confident i'm not going to mess up my setup by uploading all these old pictures.
Thanks
Beta Was this translation helpful? Give feedback.
All reactions