|Tags||kernel clustering high-availability replication|
8.4.518 Aug 2014 22:32 minor bugfix: Fix potential BUG_ON in mempool_alloc for older kernels (before 2.6.23). Fix cornercase of resync finished detection. Fix drbd_ldev_destroy to run exactly once and in worker context. Fix spurious detach/disconnect: don't announce WRITE_SAME. Fix NULL pointer deref in blk_add_request_payload (DISCARD/TRIM handling in sd). Fix regression: potential NULL pointer dereference. Fix regression: potential list corruption. Fix regression: potential deadlock in adm functions (drbdsetup). Fix regression: /sbin vs /usr/sbin in scripts. Fix race stopping the worker thread. New debugfs presence expose oldest requests and other details. Userspace moved to own repository. Linux 3.13 and 3.14 compatibility. Improve adaptive resync controler. Reduce bitmap IO during resync. Allow write-ordering policy to be bumped up again. Trigger tcp_flush_pending_frames() for PING/PING_ACK. Fix potential segfault in drbdadm adjust. Drbdsetup improvements: fix netlink socket buffers; detect modprobe failure.
ManageYou can also help out here by:
← Update project
or flagging this entry for moderator attention.