> On Sep 20, 2019, at 1:30 AM, Dimitar Maznekov <[email protected]> wrote: > > > So, even though there was no explicitly set usercache, there weren't any > > blobs in the default cache location on the old server? > > Nope. All of them are in .hg/largfiles per repo. > > > Assuming that this error is complaining about a missing largefile and > > showing the hash, can you look to see if that file is in .hg/largefiles on > > the old server? If you rsynced the repos, I don't see why it would be > > missing on the new server. > > As above - yes, file exists in sub-folders but new server complies about > missing largefiles (TortoisHG returns HTTP 500 error) > 500 is internal server error, and I don’t recall that being what is returned if a file is missing. There probably is a stacktrace of some sort in the server log. What is it? > > Have you tried `hg verify --large --lfa` on the old server? That will tell > > you if all of the blobs are on the server. If it doesn't complain about > > missing files, you might want to check .hg/largefiles again to see if > > something got added to it. > > It seams everything is ok on OLD server as: > > hg verify --large --lfa > repository uses revlog format 0 > checking changesets > checking manifests > crosschecking files in changesets and manifests > checking files > 5314 files, 315 changesets, 6544 total revisions > searching 315 changesets for largefiles > verified existence of 6 revisions of 6 largefiles > Then everything is present on the old server, and there’s no reason for it to complain about missing files. The error must be something else.
The repo on the server doesn’t have a `paths.default` config, does it?
_______________________________________________ Mercurial mailing list [email protected] https://www.mercurial-scm.org/mailman/listinfo/mercurial
