1RT-SERIALIZER(1) User Contributed Perl Documentation RT-SERIALIZER(1)
2
6 rt-serializer - Serialize an RT database to disk
7
9 rt-validator --check && rt-serializer
10 This script is used to write out the entire RT database to disk, for
12 later import into a different RT instance. It requires that the data
13 in the database be self-consistent, in order to do so; please make sure
14 that the database being exported passes validation by rt-validator
15 before attempting to use "rt-serializer".
16 While running, it will attempt to estimate the number of remaining
18 objects to be serialized; these estimates are pessimistic, and will be
19 incorrect if "--no-users", "--no-groups", or "--no-tickets" are used.
20 If the controlling terminal is large enough (more than 25 columns high)
22 and the "gnuplot" program is installed, it will also show a textual
23 graph of the queue size over time.
24 OPTIONS
26 --directory name
27 The name of the output directory to write data files to, which
28 should not exist yet; it is a fatal error if it does. Defaults to
29 "./$Organization:Date/", where $Organization is as set in
30 RT_SiteConfig.pm, and Date is today's date.
31 --force
33 Remove the output directory before starting.
34 --size megabytes
36 By default, "rt-serializer" chunks its output into data files which
37 are around 32Mb in size; this option is used to set a different
38 threshold size, in megabytes. Note that this is the threshold
39 after which it rotates to writing a new file, and is as such the
40 lower bound on the size of each output file.
41 --no-users
43 By default, all privileged users are serialized; passing
44 "--no-users" limits it to only those users which are referenced by
45 serialized tickets and history, and are thus necessary for internal
46 consistency.
47 --no-groups
49 By default, all groups are serialized; passing "--no-groups" limits
50 it to only system-internal groups, which are needed for internal
51 consistency.
52 --no-assets
54 By default, all assets are serialized; passing "--no-assets" skips
55 assets during serialization.
56 --no-disabled
58 By default, all queues, custom fields, etc, including disabled
59 ones, are serialized; passing "--no-disabled" skips such disabled
60 records during serialization.
61 --no-deleted
63 By default, all tickets and assets, including deleted ones, are
64 serialized; passing "--no-deleted" skips deleted tickets and assets
65 during serialization.
66 --scrips
68 No scrips or templates are serialized by default; this option
69 forces all scrips and templates to be serialized.
70 --acls
72 No ACLs are serialized by default; this option forces all ACLs to
73 be serialized.
74 --no-tickets
76 Skip serialization of all ticket data.
77 --limit-queues
79 Takes a list of queue IDs or names separated by commas. When
80 provided, only that set of queues (and the tickets in them) will be
81 serialized.
82 --limit-cfs
84 Takes a list of custom field IDs or names separated by commas. When
85 provided, only that set of custom fields will be serialized.
86 --hyperlink-unmigrated
88 Replace links to local records which are not being migrated with
89 hyperlinks. The hyperlinks will use the serializing RT's
90 configured URL.
91 Without this option, such links are instead dropped, and
93 transactions which had updated such links will be replaced with an
94 explanatory message.
95 --no-transactions
97 Skip serialization of all transactions on any records (not just
98 tickets).
99 --clone
101 Serializes your entire database, creating a clone. This option
102 should be used if you want to migrate your RT database from one
103 database type to another (e.g. MySQL to Postgres). It is an error
104 to combine "--clone" with any option that limits object types
105 serialized. No dependency walking is performed when cloning.
106 "rt-importer" will detect that your serialized data set was
107 generated by a clone.
108 --all
110 Serializes your entire database, creating a clone-like data. Both
111 "--all" and "--clone" do not check dependencies, the difference is
112 "--all" generates UIDs: it means the ids in source instance do not
113 necessirily be synced to target instance, which makes it quite
114 useful to fully merge multiple RT instances. Use "--clone" instead
115 if you really want to keep ids in source instance.
116 --incremental
118 Will generate an incremenal serialized dataset using the data
119 stored in your IncrementalRecords database table. This assumes
120 that you have created that table and run RT using the
121 Record_Local.pm shim as documented in "docs/incremental-export/".
122 --gc n
124 Adjust how often the garbage collection sweep is done; lower
125 numbers are more frequent. See "GARBAGE COLLECTION".
126 --page n
128 Adjust how many rows are pulled from the database in a single
129 query. Disable paging by setting this to 0. Defaults to 100.
130 Keep in mind that rows from RT's Attachments table are the limiting
132 factor when determining page size. You should likely be aiming for
133 60-75% of your total memory on an otherwise unloaded box.
134 --quiet
136 Do not show graphical progress UI.
137 --verbose
139 Do not show graphical progress UI, but rather log was each row is
140 written out.
141
143 "rt-serializer" maintains a priority queue of objects to serialize, or
144 searches which may result in objects to serialize. When inserting into
145 this queue, it does no checking if the object in question is already in
146 the queue, or if the search will contain any results. These checks are
147 done when the object reaches the front of the queue, or during periodic
148 garbage collection.
149 During periodic garbage collection, the entire queue is swept for
151 objects which have already been serialized, occur more than once in the
152 queue, and searches which contain no results in the database. This is
153 done to reduce the memory footprint of the serialization process, and
154 is triggered when enough new objects have been placed in the queue.
155 This parameter is tunable via the "--gc" parameter, which defaults to
156 running garbage collection every 5,000 objects inserted into the queue;
157 smaller numbers will result in more frequent garbage collection.
158 The default of 5,000 is roughly tuned based on a database with several
160 thousand tickets, but optimal values will vary wildly depending on
161 database configuration and size. Values as low as 25 have provided
162 speedups with smaller databases; if speed is a factor, experimenting
163 with different "--gc" values may be helpful. Note that there are
164 significant boundary condition changes in serialization rate, as the
165 queue empties and fills, causing the time estimates to be rather
166 imprecise near the start and end of the process.
167 Setting "--gc" to 0 turns off all garbage collection. Be aware that
169 this will bloat the memory usage of the serializer. Any negative value
170 for "--gc" turns off periodic garbage collection and instead objects
171 already serialized or in the queue are checked for at the time they
172 would be inserted.
173perl v5.38.0 2023-10-22 RT-SERIALIZER(1)