When the client ( a java program with jgit that runs on jboss) tries to clone a repository with files that contains about 411mb (there are files from 1 to 100 mb) it's cloning it very slow, 8 - 12 kb's second.
When i do it with git bash, it clones very fast, it take about 2 minutes to clone the repo.
With the same repository, if there is a file about 27mb it's clone it very fast. My assumption there is some trouble when a repo contains big files or the repo is more than 100mb.
Is anything that i can do to revert this situation?
JGIT Version Tested: 220.127.116.11306101825 and 18.104.22.168310021548-r
Environment: Linux Centos 6.
Try to increase jgit's window cache limit to a larger value, if there are large files to be processed
you may need to increase the stream file threshold to a value larger than the biggest file to transfer.
See parameters packedGitLimit and streamFileThreshold in class org.eclipse.jgit.storage.file.WindowCacheConfig.