1PG_RESETWAL(1) PostgreSQL 10.7 Documentation PG_RESETWAL(1)
2
6 pg_resetwal - reset the write-ahead log and other control information
7 of a PostgreSQL database cluster
8
10 pg_resetwal [-f] [-n] [option...] {[-D] datadir}
11
13 pg_resetwal clears the write-ahead log (WAL) and optionally resets some
14 other control information stored in the pg_control file. This function
15 is sometimes needed if these files have become corrupted. It should be
16 used only as a last resort, when the server will not start due to such
17 corruption.
18 After running this command, it should be possible to start the server,
20 but bear in mind that the database might contain inconsistent data due
21 to partially-committed transactions. You should immediately dump your
22 data, run initdb, and reload. After reload, check for inconsistencies
23 and repair as needed.
24 This utility can only be run by the user who installed the server,
26 because it requires read/write access to the data directory. For safety
27 reasons, you must specify the data directory on the command line.
28 pg_resetwal does not use the environment variable PGDATA.
29 If pg_resetwal complains that it cannot determine valid data for
31 pg_control, you can force it to proceed anyway by specifying the -f
32 (force) option. In this case plausible values will be substituted for
33 the missing data. Most of the fields can be expected to match, but
34 manual assistance might be needed for the next OID, next transaction ID
35 and epoch, next multitransaction ID and offset, and WAL starting
36 address fields. These fields can be set using the options discussed
37 below. If you are not able to determine correct values for all these
38 fields, -f can still be used, but the recovered database must be
39 treated with even more suspicion than usual: an immediate dump and
40 reload is imperative. Do not execute any data-modifying operations in
41 the database before you dump, as any such action is likely to make the
42 corruption worse.
43
45 -f
46 Force pg_resetwal to proceed even if it cannot determine valid data
47 for pg_control, as explained above.
48 -n
50 The -n (no operation) option instructs pg_resetwal to print the
51 values reconstructed from pg_control and values about to be
52 changed, and then exit without modifying anything. This is mainly a
53 debugging tool, but can be useful as a sanity check before allowing
54 pg_resetwal to proceed for real.
55 -V
57 --version
58 Display version information, then exit.
59 -?
61 --help
62 Show help, then exit.
63 The following options are only needed when pg_resetwal is unable to
65 determine appropriate values by reading pg_control. Safe values can be
66 determined as described below. For values that take numeric arguments,
67 hexadecimal values can be specified by using the prefix 0x.
68 -c xid,xid
70 Manually set the oldest and newest transaction IDs for which the
71 commit time can be retrieved.
72 A safe value for the oldest transaction ID for which the commit
74 time can be retrieved (first part) can be determined by looking for
75 the numerically smallest file name in the directory pg_commit_ts
76 under the data directory. Conversely, a safe value for the newest
77 transaction ID for which the commit time can be retrieved (second
78 part) can be determined by looking for the numerically greatest
79 file name in the same directory. The file names are in hexadecimal.
80 -e xid_epoch
82 Manually set the next transaction ID's epoch.
83 The transaction ID epoch is not actually stored anywhere in the
85 database except in the field that is set by pg_resetwal, so any
86 value will work so far as the database itself is concerned. You
87 might need to adjust this value to ensure that replication systems
88 such as Slony-I and Skytools work correctly — if so, an appropriate
89 value should be obtainable from the state of the downstream
90 replicated database.
91 -l walfile
93 Manually set the WAL starting address.
94 The WAL starting address should be larger than any WAL segment file
96 name currently existing in the directory pg_wal under the data
97 directory. These names are also in hexadecimal and have three
98 parts. The first part is the “timeline ID” and should usually be
99 kept the same. For example, if 00000001000000320000004A is the
100 largest entry in pg_wal, use -l 00000001000000320000004B or higher.
101 Note
103 pg_resetwal itself looks at the files in pg_wal and chooses a
104 default -l setting beyond the last existing file name.
105 Therefore, manual adjustment of -l should only be needed if you
106 are aware of WAL segment files that are not currently present
107 in pg_wal, such as entries in an offline archive; or if the
108 contents of pg_wal have been lost entirely.
109 -m mxid,mxid
111 Manually set the next and oldest multitransaction ID.
112 A safe value for the next multitransaction ID (first part) can be
114 determined by looking for the numerically largest file name in the
115 directory pg_multixact/offsets under the data directory, adding
116 one, and then multiplying by 65536 (0x10000). Conversely, a safe
117 value for the oldest multitransaction ID (second part of -m) can be
118 determined by looking for the numerically smallest file name in the
119 same directory and multiplying by 65536. The file names are in
120 hexadecimal, so the easiest way to do this is to specify the option
121 value in hexadecimal and append four zeroes.
122 -o oid
124 Manually set the next OID.
125 There is no comparably easy way to determine a next OID that's
127 beyond the largest one in the database, but fortunately it is not
128 critical to get the next-OID setting right.
129 -O mxoff
131 Manually set the next multitransaction offset.
132 A safe value can be determined by looking for the numerically
134 largest file name in the directory pg_multixact/members under the
135 data directory, adding one, and then multiplying by 52352 (0xCC80).
136 The file names are in hexadecimal. There is no simple recipe such
137 as the ones for other options of appending zeroes.
138 -x xid
140 Manually set the next transaction ID.
141 A safe value can be determined by looking for the numerically
143 largest file name in the directory pg_xact under the data
144 directory, adding one, and then multiplying by 1048576 (0x100000).
145 Note that the file names are in hexadecimal. It is usually easiest
146 to specify the option value in hexadecimal too. For example, if
147 0011 is the largest entry in pg_xact, -x 0x1200000 will work (five
148 trailing zeroes provide the proper multiplier).
149
151 This command must not be used when the server is running. pg_resetwal
152 will refuse to start up if it finds a server lock file in the data
153 directory. If the server crashed then a lock file might have been left
154 behind; in that case you can remove the lock file to allow pg_resetwal
155 to run. But before you do so, make doubly certain that there is no
156 server process still alive.
157 pg_resetwal works only with servers of the same major version.
159
161 pg_controldata(1)
162PostgreSQL 10.7 2019 PG_RESETWAL(1)