The annotated tag, ctdb-2.3 has been created at 319e230914cebb1126e469bef115c469e9d4170e (tag) tagging 412bc0e20bef694d4e911dc9c984fd7716231f1f (commit) replaces ctdb-2.2 tagged by Amitay Isaacs on Thu Jul 11 17:06:59 2013 +1000
- Log ----------------------------------------------------------------- New version 2.3 Amitay Isaacs (56): ctdbd: Make sure we don't kill init process by mistake build: Enable VERBOSE option to display build command line packaging: Update the minimum required library versions tools/ctdb: Do not exit prematurely on control timeout if retrying in a loop client: Exit with non-zero status when unix socket is closed recoverd: When updating flags on nodes, send updated flags and not old flags recoverd: Print banning message only after verifying pnn freeze: Log message from ctdb_start_freeze() and ctdb_control_freeze() freeze: If priority is invalid here, it's time to abort freeze: Make ctdb_start_freeze() a void function banning: Log ban state changes for other nodes at higher debug level recovered: Remove old comment as the code corresponding to that has gone away recoverd: Set node_flags information as soon as we get nodemap recoverd: Also check if current node is in recovery when it is banned banning: Make ctdb_local_node_got_banned() a void function banning: No need to check if banned pnn is for local node banning: Do not come out of ban if databases are not frozen recoverd: Do not set banning credits on a node if current node is inactive recoverd: Always do an early exit from main_loop if node is stopped or banned recoverd: No need to check if node is recovery master when inactive recoverd: Update capabilities only if the current node is active recoverd: Delay the initial election if node is started in stopped state recoverd: Move code to ban other nodes after we get local node flags recoverd: Refactor code to ban misbehaving nodes doc: The second half of monitoring is only for recovery master ctdbd: Don't ban self if init or shutdown event fails packaging: Remove ctdb_transaction from docdir packaging: Install docs using %doc directive packaging: Install README.notify.d using %doc directive packaging: Do not mark /etc/ctdb/functions as configuration file packaging: Allow building RPMs with system tdb/talloc/tevent packaging: Enable compiler optimizations recoverd: Send the result from child process only once build: Fix compiler warnings for uninitialized variables tests: If connection to ctdb daemon fails, exit ping_pong: Validate num_locks argument > 0 recoverd: Fix buffer overflow error in reloadips ctdbd: Update debug messages for setting readonly property on database ctdbd: Remove incomplete ctdb_db_statistics_wire structure tools/ctdb: Fix the format of DB statistics output locking: Update locks latency in CTDB statistics only for RECORD or DB locks locking: Update locking bucket intervals locking: Use external script to debug locking issues scripts: Add an example debug_locks.sh script to debug locking issue initscript: Export CTDB_DEBUG_LOCKS variable ctdbd: No need for DeadlockTimeout tunable packaging: When building with system libraries, add dependency for them traverse: Pass reqid and srcnode information to local database traverse traverse: Send records directly from traverse child to srcnode traverse: Remove unused start_time field common/system: Add ctdb_set_process_name() function ctdbd: Set process names for child processes ctdbd: Print tdb flags when logging attached to database message doc: Update NEWS Tests: Correct the arguments to memset web: Update webpages Martin Schwenke (52): recoverd: Log node that causes takoever run to fail eventscripts: Add new option $CTDB_MONITOR_NFS_THREAD_COUNT eventscripts: New configuration varable $CTDB_NFS_DUMP_STUCK_THREADS tests/eventscripts: Add unit tests for $CTDB_MONITOR_NFS_THREAD_COUNT tests/eventscripts: Fix -X tracing in iterate_test() tests/eventscripts: Unit tests for $CTDB_NFS_DUMP_STUCK_THREADS ctdbd: "init" event should run earlier in daemon initialisation scripts: drop_all_public_ips() now prints messages to stdout, not log scripts: drop_ip() should use delete_ip_from_iface() scripts: Move dropping of all IPs from initscript to "init" event scripts: Move TDB checking from initscript to "init" event logging: Notify parent when logging daemon is up tests/eventscripts: setup_ctdb() should always set $CTDB_PUBLIC_ADDRESSES eventscripts: 13.per_ip_routing should not try hard to find public_addresses tests/eventscripts: New tests for 00.ctdb "init" event eventscripts: "setup" event doesn't need to wait for SETUP runstate ctdbd: Refactor shutdown sequence ctdbd: Fix panic on overlapping shutdowns tests: Integration test infrastructure should do only a single recovery tests: Integration tests use "ctdb nodestatus" for healthy cluster check doc: Update NEWS doc: Add nodestatus command to the ctdb manpage doc: Update notification script section in ctdbd manpage doc: Fix documentation for NoIPTakeover in ctdbd manpage doc: Fix ctdb ping entry in manpage tools/ctdb: Add "force" option to "recover" command recoverd: Don't continue if the current node gets banned doc: Add banning bug fixes to NEWS doc: Add a disclaimer for the EnableBans tunable ctdbd: Log warnings in release IP when unexpected interface is encountered ctdbd: Release IP callback should fail if the IP is still hosted eventscripts: When replaying monitor status, don't log empty output util: New function ctdb_die() ctdbd: Avoid a core dump when "init" event fails ctdbd: Use ctdb_die() on "setup" event failure recoverd: Fix an incorrect comment recoverd: Fix an unclear log message - "Restart recovery process" recoverd: Clean up log messages in remote IP verification recoverd: Minor style improvements for ctdb_reload_remote_public_ips() ctdbd: Log something when releasing all IPs eventscripts: Drop RPC service version from nfs_check_rpc_service() calls eventscript: Move Ganesha nfsd monitoring to a function eventscripts: New configuration variable $CTDB_SKIP_GANESHA_NFSD_CHECK tests/eventscripts: Add some rudimentary tests for 60.ganesha recoverd: Recovery daemon should use ctdb_get_pnn, which can't fail initscript: Simpify initscript and control CTDB via new ctdbd_wrapper build: Remove -DTEVENT_DEPRECATED_QUIET=1 from CFLAGS build: Turn off all deprecation warnings packaging: Add systemd support doc: Update NEWS recoverd: Really fix bogus info in message about changed flags scripts: ctdbd_wrapper logs a message to syslog if syslog is not being used Mathieu Parent (3): libctdb: Include config.h in libctdb/ctdb.c build: Fix tdb.h path to enable building with system TDB library Update Nagios check to work with ctdb versions past 30 Aug 2011 Michael Adam (4): recoverd: eliminate some trailing spaces from ctdb_election_win() recoverd: fix a comment in main_loop recoverd: fix a comment typo recoverd: when the recmaster is banned, use that information when forcing an election Sumit Bose (5): Check return value of tdb_delete() Fixes for various issues found by Coverity Fix memory leak in ctdb_send_message() IPv6 neighbor solicit cleanup Print deleted nodes as well ----------------------------------------------------------------------- -- CTDB repository