summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorKent Overstreet <kent.overstreet@gmail.com>2016-08-11 07:11:07 -0800
committerKent Overstreet <kent.overstreet@gmail.com>2016-08-11 07:11:07 -0800
commit8e0e14f5b8e7caba2522e3c8ad15bef6a282329a (patch)
treeebe8b0ce68d7ea4897f76743d813121a49d433a6
parent5bf422eb65b481f8fa43f05af8d87a1db54e808e (diff)
update bcachefs status
-rw-r--r--index.mdwn16
1 files changed, 12 insertions, 4 deletions
diff --git a/index.mdwn b/index.mdwn
index 69c072b..8fe8b94 100644
--- a/index.mdwn
+++ b/index.mdwn
@@ -21,10 +21,18 @@ kernel, and a long list of features - among them, multiple devices, replication,
caching and full data checksumming/compression. Snapshotting is coming, too.
Pretty much all the normal posix filesystem stuff is supported (things like
-xattrs, acls, etc. - no fallocate or quotas yet, though).
-
-Be warned - the on disk format has not been frozen yet, and it might eat all
-your data (but it is passing all but a few of the tests in xfstests).
+xattrs, acls, etc. - no quotas yet, though).
+
+Stability: there are no known outstanding bugs if you stick to the single device
+filesystems (no tiering/replication). It's been passing xfstests for months, and
+it is seeing real world usage (I've been using it on my development laptop for
+some months now). It is highly unlikely to eat your data, but that is always a
+possibility with new filesystems - please keep backups.
+
+The on disk format is not yet set in stone, but any future breaking changes will
+come with plenty of warning. We'll need at least one more breaking change for
+encryption and possibly snapshots, but I'm hoping to delay rolling out breaking
+changes until I can do them at once.
For other kernel programmers who might be interested in getting involved, I've
started a guide to the bcache internals: [[BcacheGuide]]