Changelog History
Page 1
-
v3.0.115 Changes
October 09, 2020Dear MooseFS Users,
We are happy to announce that MooseFS 3.0.115 is now stable!
๐ This is a bugfix release. We would like to say big thanks to all the contributors who helped us making this release, especially (alphabetically): @borkd, @eleaner, @richarson. Thank you for your valuable time, input, and help in making MooseFS better!
(If I forgot to mention somebody, I am sorry and please let me know!)
๐ Please find a complete list of fixes and new features that this release includes, below:
- ๐ (cs) fixed chunk locking issues in chunkserver (lock timeout, missing release etc.)
- (all) changed to hex all
chunkid
logged to syslog (only in changelog it is left in decimal format) - (cs) changed port number from
hex
todec
in replicator module - ๐จ (mount) added printing
open
/create
mode to the.oplog
- ๐ (mount) fixed opening file after using
lookup
cache - (master) added ignoring empty filenames
โก๏ธ We recommend to update your MooseFS cluster(s) to this version.
Best regards,
Piotr / MooseFS Team -
v3.0.114 Changes
July 29, 2020Dear MooseFS Users,
We are happy to announce that MooseFS 3.0.114 is now stable!
๐ This is mostly a bugfix release, however it also introduces a few improvements, including:
- ๐ Improved internal rebalance algorithm in Chunkserver,
- More information reported in
.params
special file on MooseFS Client side, - ๐ Changes in job queue policy in Chunkserver,
- ๐ Changed time format in MooseFS CGI Monitoring Interface,
- ๐ Improvements related to chunks database file (
.chunkdb
) on Chunkserver (logging, write error detection).
๐ We would like to say big thanks to all the contributors who helped us making this release, especially (alphabetically):
๐ท @BillKen, @jSML4ThWwBID69YC, @kopaczew, @onlyjob, @uu.
๐ Thank you for your valuable time, input, and help with making MooseFS better!(If I forgot to mention somebody, I am sorry and please let me know!)
๐ Please find a complete list of fixes and new features that this release includes, below:
- ๐ (mfsio/nbd) fixed compile issue on systems without
LOCK_*
constants (issue #366) - ๐ (mount) fixed truncating files open with
O_APPEND
flag (bug introduced in 3.0.113, issue #368) - (cs) added disk rebalance when all designated source disks are already empty (issue #364)
- (mount) added additional info to
.params
file (versions of MooseFS,libfuse
etc.) - (master+cs) added check against
EACCES
afterlockf
(issue #369) - ๐ (master) fixed access checks in snapshot
- ๐ (cgi+cli) removed using supervisor (only available in Pro version)
- ๐ (cs) fixed deadlock condition on mutexes
folderlock
andhashlock
(exists since 3.0.110) - ๐ (mount) fixed handling
.params
file - ๐ (tools) fixed overflow protection in number parsing function
- ๐ (cs) fixed reporting to master duplicates with newer version
- ๐ท (cs) changed job queue policy (some tasks are not limited now)
- ๐ (cli) fixed help and manpage
- ๐ (master) fixed reaction to status
NOTDONE
received from chunkservers - ๐ (master) fixed syncing and closing negative file descriptor in
bgsaver
- ๐ (mount+master) fixed handling
keepcache
anddirect
flags (related to issue #374) - ๐ (mount) fixed compiling on systems with old
fuse
library - (cgi) changed time format (issue #197)
- ๐ฒ (cs) added logging info when
.chunkdb
is not written to disk - ๐ (master) fixed alphabetical order of commands detected in changelog
- ๐ท (cs) changed queue limit to max workers (limited dynamically)
- ๐ (cs) fixed automatic chunkserver removal in master
- (master) added meta version increment in
chunks_set_version
- (cs) added error detection during writing
.chunkdb
- ๐ (all) added build id to
what
strings - (all) added missing
idstr.h
and rearranging source lists in Makefiles - (nbd+cgiserv) added
chdir("/")
- โก๏ธ (freebsd) updated
rc.d
master script - โก๏ธ (man) updated mfsmetadump manpage
- ๐ (man) fixed formatting in mfsmount manpage
-
v3.0.113 Changes
May 12, 2020Dear MooseFS Users,
We are happy to announce that MooseFS 3.0.113 is now stable!
๐ This is both bugfix and feature release. Apart from fixing the issues that were recently found, it introduces:
- ๐ Support for new extra attributes:
immutable
,appendonly
,undeletable
.
๐ Extra attributes are MooseFS-specific feature that allows you to add / manage a small portion of metadata to files on MooseFS. There are several extra attributes and they change file system behaviour (apart from new ones mentioned above there are several more โ please refer toman mfseattr
for more information). - Topology grouping when new chunks are about to be created,
- Readonly mode and locking in MooseFS Block Device,
- Changing subfolder during internal rebalance,
- ๐ง Official labels format defined in chunkserver configuration,
and several more features.
๐ This release improves Chunkserver a lot (registering chunks to Master Server, reporting damaged chunks, internal rebalance, reloading, calculating size limits, handling inode limits on underlying file systems). It also adds new and improves existing workarounds to issues related to OS's kernels and improves directory entry invalidator introduced in MooseFS 3.0.112.
๐ We would like to thank all the contributors who helped us making this release, especially (alphabetically):
๐ท @borkd, @eleaner, @jSML4ThWwBID69YC, @onlyjob, @rafamiga, @tamas-ac.
๐ Thank you for your valuable time, input, and help with making MooseFS better!(If I forgot to mention somebody, I am sorry and please let me know!)
๐ Please find a complete list of fixes and new features that this release includes, below:
- ๐ (master) removed unnecessary debug syslog messages
- (check) increased usleep tolerance due to some operating systems (issue #351)
- (master) changed condition that decides if master should wait for more chunks during I/O
- (mount) delayed setting channel for fuse notifications (very rare segfault in
libfuse2
during init) - โ
(client+master) changed conditions for ancestor test in
getattr
(issue #350) - (cs) added parameter for number of chunks to be send in single register packet
- โ (cs) fixed reporting damaged chunks in testing function (issue #352)
- ๐ (cs) removed some unnecessary damaged chunk notifications
- โ
(client) changed open test in
setattr
(related to issue #350) - ๐ (master) fixed handling truncate for open files (related to issue #350)
- (master) added uid mapping in
setfacl
function - ๐ (master+client) added support for atomic truncate with open
- ๐ (master+client) fixed keep cache conditions
- ๐ง (cs) introduced official label format defined in chunkserver configuration
- (nbd) added readonly mode and locking (issue #353)
- ๐ (cs) fixed condition in choosing disks for internal rebalance
- โช (mount) added workaround in access for a bug in FreeBSD kernel (issue #354)
- ๐ (cs) fixed master reconnection conditions (reload usually shouldn't cause reconnection)
- ๐ (freebsd) fixed FreeBSD port (
makeports.sh
) - (master) changed disk removal detection algorithm (issue #356)
- ๐ (cs) fixed calculating size limits
- (cs) added handling inode limits on local filesystems on chunkservers (issue #358)
- (master) added topology grouping when new chunks are about to be created (prefer closer servers)
- (master) added more error messages to
bgsaver
- (cs) added changing subfolder during internal rebalance (related to issue #326)
- ๐ฒ (master+cs) added optional logging of long function execution times
- ๐ (all) fixes of small errors found by static code analysers
- ๐ง (mount) turn off
dentry
invalidator for Linux kernels >= 4.19 (related to issue #357) - ๐ (master) changed condition that checks timestamp in changelogs (less prone to small clock differences)
- (daemons) added time refresh function (for future use)
- (cgi+cli) changed mark for removal state name
UNKNOWN
โ>PENDING
(related to issue #359) - (client) decreased max
IDLE
time in writer module from 1s to 0.1s - โก๏ธ (client) update inode in
dentry
invalidator - ๐ (tools) fixed
mfscopyeattr
,mfsseteattr
andmfs*archive
tools - ๐ (master+client) added support for new
eattrs
:immutable
,appendonly
,undeletable
(setting newundeletable
extra attribute appropriately can help with issues like #357) - (client) silenced
kern.proc.filedesc
syslog messages on FreeBSD (issue #360)
Best regards,
Piotr / MooseFS Team - ๐ Support for new extra attributes:
-
v3.0.112 Changes
March 24, 2020Dear MooseFS Users,
MooseFS 3.0.112 is now stable!
๐ This release is a bugfix and feature release. It introduces support for disabling individual filesystem commands in Master Server (configurable in
mfsexports.cfg
) and respecting them on the Client side (issue #344). It also adds support forSTRICT
andLOOSE
mode inKEEP
andARCHIVE
(issue #249). This release fixes various other issues too, including a workaround for Linux kernel bug related toEBUSY
issue by adding directory entry invalidator (issue #98).๐ We would like to thank all the contributors who helped us making this release, especially (alphabetically):
๐ง @borkd, @Damien9527, @linux-ops, @MonkeyFang, @onlyjob, @pault28, @tnktls.
๐ Thank you for all your valuable input, help and making MooseFS better!(If I forgot to mention somebody, I am very sorry and please let me know!)
๐ Please find a complete list of fixes and new features that this release includes below:
- โ (cs) silence stupid compiler warning
- ๐ (client) fixed handling
LOCKED
andEAGAIN
status inreaddata
- (client) added handle info to
.oplog
messages - ๐ (all) added support for disabling individual filesystem commands in
mfsexports.cfg
- ๐ (client) added handling
read
/write
/readdir
disables on client side (better error messages) - (client) added session paramaters to
.params
file - ๐ (tools) fixed packet reallocation error
- (master) changed
syslog
message for locked chunks - ๐ง (client) added
dentry
invalidator (needed in Linux with kernel < 4.19 โEBUSY
issue) - ๐ (master) fixed memory leaks in
xattr
andposixacl
modules - (cs) when chunk can't be located always send to master info about lost chunk
- ๐ (supervisor, master, cs) fixed buffer overrun in
mastersupervisor
code (HA only, intr. in 3.0.107) - ๐ (master) changed algorithm of reusing
csid
in chunk module (adding releasedcsid
to the end of free list) - ๐ (all) added support for
STRICT
andLOOSE
mode inKEEP
andARCHIVE
- ๐ (master) fixed segfault during appending chunks of file with positive length and no chunks
- ๐ (metadump) fixed bugs introduced in version 3.0.106
- (master) added protection between never-ending desyncing between newer LEADER and older FOLLOWER (HA only)
- ๐ (master) fixed slices with
to
set to 0 inmfsappendchunks
- ๐ (master) removed starting protection time from client communication module
- ๐ (man) fixed typo in
mfsappendchunks
man page
Best regards,
Piotr / MooseFS Team -
v3.0.111 Changes
February 19, 2020Dear MooseFS Users,
๐ MooseFS 3.0.111 is now stable! This release adds protections against segfaults in MooseFS Chunkserver module (issue #342).
๐ท We would like to thank @onlyjob and @uu for reporting the issue. Thank you for all your valuable input, help and making MooseFS better!
๐ Please find a complete list of fixes and new features that this release includes below:
- (cs) added protections against segfaults (issue #342)
- ๐ (cs) removed unnecessary error messages
Best regards,
Piotr / MooseFS Team -
v3.0.110 Changes
February 13, 2020Dear MooseFS Users,
๐ MooseFS 3.0.110 is now stable! This release is a bugfix release mainly. It addresses issues related to handling duplicated chunks, placing chunks in subfolders on Chunkservers, internal rebalance, fixes FreeBSD-related issues and adds a few other minor fixes.
๐ This release also introduces several new config options, new logo and favicon in MFS CGI interface and possibility to dump non-printable characters in metadata dumper.
๐ We would like to thank all the contributors who helped us making this release, especially (alphabetically): @borkd, @jSML4ThWwBID69YC, @onlyjob and @smurfix. Thank you for all your valuable input, help and making MooseFS better!
(If I forgot to mention somebody, I am very sorry and please let me know!)
๐ Please find a complete list of fixes and new features that this release includes below:
- ๐ (cs) fixed bug that may lead to creating much more chunks in one subfolder (issues #319, #326)
- (cs) added option for controlling subdirectory selection algorithm for newly created chunks (issues #319, #326)
- (cs) prevent potential dumping deleted chunks to
.chunkdb
- (client) added options for setting behaviour when master returns unavailable chunks or no space status (issue #322)
- ๐ (cgi) fixed logo and icon (file type for
favicon.ico
, change to current logotype โ issue #328) - (master) turning off sending NOPs to unregistered clients (
matoml
module) - ๐ (cs) fixed reporting wrong number of hours in duplicates removal message
- (cs) redesigned duplicates removal procedure to do it in groups of 100 files (issue #327)
- (cs) introduce
mypread
andmypwrite
macro (code cleaning) - (cs) added check for changed CRC before every chunk file close (just in case)
- (cs) never use open chunk as a chunk duplicate that needs to be deleted
- (cs) never use disk with pending duplicates as destination disk during internal rebalance (issue #327)
- ๐ (cs) fixed condition in internal rebalance procedure (issue #327)
- (cs) added option for setting duplicates retention time in hours (issue #333)
- ๐ (cs) fixed memory leak in removing disk with duplicates
- ๐ (cs) removed unnecessary attribute reading when chunk duplicates are detected (issue #327)
- (cs) changed allocation of duplicated chunks from malloc to mmap, decreased memory usage by duplicates
- (client) reading in
O_WRONLY
no longer causesEACCES
(might be related to issue #334) - ๐จ (metadump) printing dots for non-printable characters made optional (issue #332)
Best regards,
Piotr / MooseFS Team -
v3.0.109 Changes
November 30, 2019Dear MooseFS Users,
๐ MooseFS 3.0.109 is now stable! This release fixes one minor bug in MooseFS CGI Server introduced in 3.0.108 and adds a possibility to allow starting MooseFS Chunkserver with invalid entries in
mfshdd.cfg
.๐ We would like to thank all the contributors who helped us making this release. Thank you for all your help!
๐ Please find the list of fixes and new features that this release includes below:
- ๐ (mfscgiserv) fixed handling redirects (bug introduced in 3.0.108)
- (cs) added allowance for starting with invalid entries in
mfshdd.cfg
Best regards,
Piotr / MooseFS Team -
v3.0.108 Changes
November 30, 2019Dear MooseFS Users,
๐ MooseFS 3.0.108 is now stable! This release is a bugfix release.๐ We would like to thank all the contributors who helped us making this release, especially (alphabetically): @onlyjob, @Programie, @xandrus. Thank you for all your help!
(If I forgot to mention somebody, I am very sorry and please let me know!)
๐ Please find the list of fixes that this release includes below:
- ๐ (cli+cgi) fixed bug related to
masterinfolist
- ๐ (master) fixed sign in reporting time difference
- (cgiserv) added
content-type: text/plain
to cgi outputs with errors - ๐ (master) fixed
rmdir
packet size
Best regards,
Piotr / MooseFS Team - ๐ (cli+cgi) fixed bug related to
-
v3.0.107 Changes
November 11, 2019Dear MooseFS Users,
Long-awaited MooseFS 3.0.107 is now stable!This release is mainly a bugfix release (and includes a lot of fixes). We strongly recommend to upgrade to this latest version! This release also introduces some new features, including support for
libfuse3
.๐ We would like to thank all the contributors who helped us making this release, especially (alphabetically): @eleaner, @hradec, @itongqingbo, @njaard, @onlyjob, @steffenb7333. Thank you for all your help!
(If I forgot to mention somebody, I am very sorry and please let me know!)
๐ Please find the list of fixes and new features that this release includes below:
- (master) added NOP command to changelog
- (master+supervisor) added time check
- ๐ท (cs) fixed behaviour when job queue is full
- (cli+cgi) added time data to master servers
- ๐ (cs) fixed NOP packets ping-pong
- ๐ (client) fixed NOP flooding in write module
- ๐ท (cs) fixed race condition in
bgjobs
module (jobhash
) - ๐ (cs+mount) added new status returned to mount when queue is full and support for it in client module
- ๐ (mount) added initial support for
libfuse3
(related to issue #302) - ๐ (mount) added support for
readdirplus
(related to issue #302) - โช (mount) workaround for bug in
fuse
(wronglock_owner
in flush โ issue #305) - ๐ (mount) fixed deadlock in
readdata.c
(closes issue #306) - โ (mount+master) fixed few small issues related to POSIX compliance (checked by pjd fstest)
- โ (all) changed preferred C standard from C99 to C11 and added tests for atomics
- ๐ (mount) move read/write structures initialization from open to read/write
- (mount,mfsio) added new
inode->fleng
map - ๐ง (configure) added option
disable_all
, added summary at the end - (freebsd) detecting new
fuse
version in the kernel and adjusting behaviour - ๐ (tools) better handling of unrecognized options
- ๐ (master) better hard links handling
- (master) added limit for number of hardlinks allowed for one object
- (master) added option for automatic removal of unused chunkservers (issue #247)
- ๐ (metadirinfo) fixed segfault when output filename is not specified
- (chunktool) added file name fixing option (issue #276)
- (all) changed chunk replication and deletion charts into two-color charts (successful/unsuccessful โ issue #273)
- (all) added new charts with number of chunk creations and internal changes (also successful and unsuccessful)
- (master) added ignoring chunk with id zero sent from chunkserver
- ๐ (metadump) added option for dumping parts of metadata
- ๐ (master) added more logs to background data write subprocess (related to issue #280)
- ๐ (master) fixed LEADER/FOLLOWER desync on SETCHUNK command (pro only)
- ๐ง (cli+cgi) fixed shebang in
python
scripts (preference forpython3
in configure) - (master) changed pipe in
bgsaver
to nonblocking mode - (master) added check if the background saving process is alive
- (master) added lockfile for background saving process
- ๐ (master) added protection against time going back in the free inodes list
- (master) added protection in FOLLOWER against not keeping up with the LEADER (pro only)
- (master) changed condition for removing chunk from data structures (all chunkservers have to be connected)
- โช (master) added
restore
run mode (works similar tomfsmaster -a
but without running daemon) - (master) take write replication counters into account in server ordering for new chunk creations
- (master) changed calculating replication limits in chunk rebalance procedure (using actual usage difference)
Best regards,
Piotr / MooseFS Team -
v3.0.105 Changes
May 27, 2019๐ MooseFS 3.0.105 is now stable! This release is mainly a bugfix release, however it also introduces a new feature of specifying subfolder for Block Devices on MooseFS.
๐ We would like to thank all the contributors who helped us making this release, especially (alphabetically): @dragon2611, @eleaner, @jkiebzak, @onlyjob, @xandrus. Thank you for all your help!
(If I forgot to mention somebody, I am very sorry and please let me know!)
๐ Please find the list of fixes this release includes below:
- ๐ (nbd) fixed segfault when filename not specified
- ๐ (client) fixed size of waiting threads counter (very rare deadlock in write module)
- ๐ (cli) fixed status returned in case of master name that can't be resolved (issue #238)
- (master) improved readability of some functions (related to issue #241)
- (nbd) added option for specifying subfolder on MFS for block devices (issue #252)
- ๐ (netdump) fixed using deprecated function
pcap_lookupdev
- ๐ (cli) fixed 'free space' in plain mode
Best regards,
Piotr / MooseFS Team