Reiterate that 'file cache names are absolute'
Drive home the point that relative source names does not save you from re-chunking if absolute pathnames change.
Also: mention bind mount solution.
remove WSL autodetection. if WSL still has this problem, you need to
set BORG_WORKAROUNDS=basesyncfile in the borg process environment to
work around it.
(cherry picked from commit beb948fc71)
docs about remote temp directory usage
Updated doc to cover use of temp directory on remote, in particular how to handle changing the directory used for systems such as QNAP with tiny default /tmp directories.
fixup by tw: do not mention specific brands, specific amounts
this is not QNAP specific, but may happen as well on other limited
systems. the amount of tmp files can also vary widely.
Fix various code blocks in the docs
- rst markup: put codeblock markup on separate line to make it better visible / separate it from normal text colons.
- borg help texts in archiver.py: put codeblock markup instead of colon - that way it looks like a single colon when using the cli help and also works as code block markup.
add a push-style example to borg-create docs
When I browsed the readthedocs and manpages while setting up Borg for
myself for the first time, I saw multiple hints that pushing backups via
SSH is supported (e.g. the existence of the `borg serve` command and the
`--rsh` option), but I never found a clear example. I only found
<https://borgbackup.readthedocs.io/en/stable/usage/general.html#repository-urls>
after being pointed to it by @ThomasWaldmann.
Therefore, an explicit example in the `borg create` documentation seems
appropriate. I put it next to the pull-style example using sshfs which
serves a similar usecase.
It was broken because there's already a different reference named
github in `docs/global.rst.inc`. Fixed by using an anonymous hyperlink
reference, per https://stackoverflow.com/a/14067756/278488
Add "SSH Configuration" section
Add "SSH Configuration" section to "borg serve" documentation, to outline ssh/sshd configuration to prevent borg serve keeping a lock on a repo in the event the ssh connection is abnormally disconnected.
In response to issues #3988, #636 and #4485 (and probably others).
drop BORG_HOSTNAME_IS_UNIQUE (please use BORG_HOST_ID if needed)
borg now always assumes it has a unique hostid - either automatically
from fqdn plus uuid.getnode() or overridden via BORG_HOST_ID.
