0
I can a server with git (home made) "private" repositories. I’m having problems with one of the repositories. It turns out that this Repo has already reached the mark of 15GB on server. When I try to make a git clone of it I get the message from:
error: pack-objects died of signal 9828/1835
error: git upload-pack: git-pack-objects died with error.
fatal: git upload-pack: aborting due to possible repostiory corruption on the remote side.
remote: aborting due to possible repository corruption on the remote site.
fatal: early EOF
fatal: index-pack failed
As I understand it, this message is generic, because it used to have the repositories in a Raspberry pi. It gave problems of memory. Okay, I switched to an ATOM server with 1 GB of ram running Ubuntu server. At first commit went well, however Repo started to increase and again this error message.
I tried to do the following:
Copy the Repo files on one of the client PC and start a new Repo.
Turns out before you do "git add ." and "git commit" the folder had about 8GB. After these two commands the folder got 15Gb. Ok, I realize it is due to the fact of the files Hidden that git itself creates in the Repo folder for management.
The point is that even after doing the "first commit" in this new I tried to do the "git push origin master" and stopped at 39%. After 2 hours so I canceled the process.
So we have two problems:
- The file on the server is no longer able to make git clone of dude!
- Tried to create a new Reset with the old files (without Hidden), thus losing all commit from the first Repo, no I can push to server;
Note: In the original client-side version, I can make git push origin master, it takes a while but it passes! It turns out I only have Repo on a client! The rest of them don’t have access to Repo because it’s not possible to make git clone of it.
I’ve been searching, and I found this topic (Handle large files with GIT) who speaks of this tool: http://git-annex.branchable.com/
Right now I don’t know how to solve this problem. I appreciate the help.
It’s worth a read: https://git-lfs.github.com/
– gmsantos
But is there a solution? Or will I have to build a new environment? I am very urgent, since I only have a valid copy of the system!
– Edgar Oliveira
As I understand it is installed on the client side, right? On the site it says to do "git lsf track "*. psd", to track all psd. But in my case I intend to do for a complete repository. How should I do?
– Edgar Oliveira
Why is this Reset so big? Are these binary files? Media? What type?
– egomesbrandao
@egomesbrandao Are pdf’s, mp4, and words!
– Edgar Oliveira
Can you talk a little bit more about the need to have these files in the repository? They’re really important there?
– egomesbrandao
@egomesbrandao São! It is a repository that has to be shared by 4 people. We all have to work on various excels or words and consequently make pdfs of the same! At the end of two months of work the Repo is with a size of +/- 10Gb!
– Edgar Oliveira
@Edgaroliveira would love to understand this need... Of you to have in the middle of 10Gb Word, Excel, PDF development process...
– egomesbrandao
@selfsame to the main question here is how do I make git manage all this!? From what I’ve read I’ll have to use LFS. You know how to help me with that?
– Edgar Oliveira
@Edgaroliveira, the issue is not only this, I’m willing to understand your process and need to recommend you the best solution. I’ve seen people storing thousands of PDF files in a repository without any need, just needed to change the process, since PDF could be generated again with data in the database and had no reason to exist physically. At this time see if this helps you to get started https://www.youtube.com/watch?v=9gaTargV5BY
– egomesbrandao