1MFSCHUNKSERVER.CFG(5) MFSCHUNKSERVER.CFG(5)
2
6 mfschunkserver.cfg - main configuration file for *mfschunkserver*
7
9 The file mfschunkserver.cfg contains configuration of LizardFS
10 chunkserver process.
11
13 Syntax is:
14 OPTION = VALUE
16 Lines starting with # character are ignored.
17
19 Configuration options:
20 DATA_PATH
22 where to store files with usage statistics and daemon lock file
23 LABEL
25 the label of this chunkserver (for tiering)
26 WORKING_USER
28 user to run daemon as
29 WORKING_GROUP
31 group to run daemon as (optional - if empty then default user group
32 will be used)
33 SYSLOG_IDENT
35 name of process to place in syslog messages (default is
36 mfschunkserver)
37 LOCK_MEMORY
39 whether to perform mlockall() to avoid swapping out mfschunkserver
40 process (default is 0, i.e. no)
41 NICE_LEVEL
43 nice level to run daemon with (default is -19 if possible; note:
44 process must be started as root to increase priority)
45 MASTER_HOST
47 address of LizardFS master host to connect with (default is
48 mfsmaster)
49 MASTER_PORT
51 number of LizardFS master port to connect with (default is 9420)
52 MASTER_RECONNECTION_DELAY
54 delay in seconds before trying to reconnect to the master server
55 after disconnection (default is 5)
56 MASTER_TIMEOUT
58 timeout (in seconds) for the master server connection (default is
59 60, minimum is 0.01)
60 BIND_HOST
62 local address to use for connecting with the master server (default
63 is *, i.e. default local address)
64 CSSERV_LISTEN_HOST
66 IP address to listen on for client (mount) connections (* means
67 any)
68 CSSERV_LISTEN_PORT
70 port to listen on for client (mount) connections (default is 9422)
71 CSSERV_TIMEOUT
73 timeout (in seconds) for client (mount) connections (default is 5)
74 HDD_CONF_FILENAME
76 alternative name of mfshdd.cfg file
77 HDD_LEAVE_SPACE_DEFAULT
79 free space threshold to set volume as 100% utilized when there is
80 less than given amount of free space left (default is "4GiB"). This
81 number is always added to the used disk space reported by
82 chunkserver.
83 HDD_TEST_FREQ
85 chunk test period in seconds (default is 10)
86 HDD_ADVISE_NO_CACHE
88 whether to remove each chunk from page when closing it to reduce
89 cache pressure generated by chunkserver (default is 0, i.e. no)
90 HDD_PUNCH_HOLES
92 if enabled then chunkserver detects zero values in chunk data and
93 frees corresponding file blocks (decreasing file system usage).
94 This option works only on Linux with file systems supporting
95 punching holes (XFS, ext4, Btrfs, tmpfs)
96 ENABLE_LOAD_FACTOR
98 if enabled, chunkserver will send periodical reports of its I/O
99 load to master, which will be taken into consideration when picking
100 chunkservers for I/O operations.
101 REPLICATION_BANDWIDTH_LIMIT_KBPS
103 limit how many kilobytes can be replicated from other chunkservers
104 to this chunkserver in every second (by default undefined, i.e. no
105 limits)
106 NR_OF_NETWORK_WORKERS
108 number of threads which handle (in a round-robin manner)
109 connections with clients (default is 1); these threads are
110 responsible for reading from sockets and coping data from internal
111 buffers to sockets
112 NR_OF_HDD_WORKERS_PER_NETWORK_WORKER
114 number of threads that each network worker may use to do disk
115 operations like opening chunks, reading or writing them (default is
116 2)
117 READ_AHEAD_KB
119 additional number of kilobytes which should be passed to
120 posix_fadvise(POSIX_FADV_WILLNEED) before reading data from a chunk
121 (default is 0, i.e. use posix_fadvise only with the amount of data
122 that is really needed; the value is aligned down to 64 KiB)
123 MAX_READ_BEHIND_KB
125 try to fix out-of-order read requests; the value tells how much of
126 skipped data to read if an offset of some read operation is greater
127 than the offset where the previos operation finished (default is 0,
128 i.e. don’t read any skipped data; the value is aligned down to 64
129 KiB)
130 CREATE_NEW_CHUNKS_IN_MOOSEFS_FORMAT
132 whether to create new chunks in the MooseFS format (signature +
133 <checksum>* + <data block>*) or in the newer interleaved format
134 ([<checksum> <data block>]*). (Default is 1, i.e. new chunks are
135 created in MooseFS format)
136 PERFORM_FSYNC
138 call fsync() after a chunk is modified (default is 1, i.e. enabled)
139 REPLICATION_TOTAL_TIMEOUT_MS
141 total timeout for single replication operation. Replications that
142 take longer than that are considered failed and are immediately
143 aborted (default: 60000)
144 REPLICATION_CONNECTION_TIMEOUT_MS
146 connection for single replication operation. If connecting to
147 another chunkserver takes longer than this timeout, the operation
148 is considered failed and is immediately aborted (default: 1000)
149 REPLICATION_WAVE_TIMEOUT_MS, ADVANCED
151 timeout for single wave in replication. After this timeout, next
152 wave of read requests is sent to other chunkservers (default: 500)
153
155 Copyright 2008-2009 Gemius SA, 2013-2015 Skytechnology sp. z o.o.
156 LizardFS is free software: you can redistribute it and/or modify it
158 under the terms of the GNU General Public License as published by the
159 Free Software Foundation, version 3.
160 LizardFS is distributed in the hope that it will be useful, but WITHOUT
162 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
163 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
164 for more details.
165 You should have received a copy of the GNU General Public License along
167 with LizardFS. If not, see http://www.gnu.org/licenses/.
168
170 mfschunkserver(8), mfshdd.cfg(5)
171 07/20/2023 MFSCHUNKSERVER.CFG(5)