store_api/
mito_engine_options.rs

1// Copyright 2023 Greptime Team
2//
3// Licensed under the Apache License, Version 2.0 (the "License");
4// you may not use this file except in compliance with the License.
5// You may obtain a copy of the License at
6//
7//     http://www.apache.org/licenses/LICENSE-2.0
8//
9// Unless required by applicable law or agreed to in writing, software
10// distributed under the License is distributed on an "AS IS" BASIS,
11// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12// See the License for the specific language governing permissions and
13// limitations under the License.
14
15//! Option keys for the mito engine.
16//! We define them in this mod so the create parser can use it to validate table options.
17
18use std::collections::HashMap;
19
20/// Option key for all WAL options.
21pub use common_wal::options::WAL_OPTIONS_KEY;
22/// Option key for append mode.
23pub const APPEND_MODE_KEY: &str = "append_mode";
24/// Option key for the per-region write buffer size.
25pub const WRITE_BUFFER_SIZE_KEY: &str = "write_buffer_size";
26/// Option key for merge mode.
27pub const MERGE_MODE_KEY: &str = "merge_mode";
28/// Option key for TTL(time-to-live)
29pub const TTL_KEY: &str = "ttl";
30/// Option key for the per-table auto flush interval.
31pub const AUTO_FLUSH_INTERVAL_KEY: &str = "auto_flush_interval";
32/// Option key for compaction type.
33pub const COMPACTION_TYPE: &str = "compaction.type";
34/// Option key for forcing compaction options override.
35pub const COMPACTION_OVERRIDE: &str = "compaction.override";
36/// TWCS compaction strategy.
37pub const COMPACTION_TYPE_TWCS: &str = "twcs";
38/// Option key for twcs min file num to trigger a compaction.
39pub const TWCS_TRIGGER_FILE_NUM: &str = "compaction.twcs.trigger_file_num";
40/// Option key for twcs min file num to trigger compaction in the active window.
41pub const TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM: &str =
42    "compaction.twcs.active_window.trigger_file_num";
43/// Option key for the active-window L1 safety compaction threshold.
44pub const TWCS_ACTIVE_WINDOW_L1_MERGE_TRIGGER: &str =
45    "compaction.twcs.active_window.l1_merge_trigger";
46/// Option key for twcs min file num to trigger compaction in an inactive window.
47pub const TWCS_INACTIVE_WINDOW_TRIGGER_FILE_NUM: &str =
48    "compaction.twcs.inactive_window.trigger_file_num";
49/// Option key for the inactive-window L1 compaction threshold.
50pub const TWCS_INACTIVE_WINDOW_L1_MERGE_TRIGGER: &str =
51    "compaction.twcs.inactive_window.l1_merge_trigger";
52/// Option key for twcs max output file size.
53pub const TWCS_MAX_OUTPUT_FILE_SIZE: &str = "compaction.twcs.max_output_file_size";
54/// Option key for twcs time window.
55pub const TWCS_TIME_WINDOW: &str = "compaction.twcs.time_window";
56/// Option key for twcs remote compaction.
57pub const TWCS_REMOTE_COMPACTION: &str = "compaction.twcs.remote_compaction";
58/// Option key for twcs fallback to local.
59pub const TWCS_FALLBACK_TO_LOCAL: &str = "compaction.twcs.fallback_to_local";
60/// Option key for memtable type.
61pub const MEMTABLE_TYPE: &str = "memtable.type";
62/// Option key for bulk memtable merge threshold.
63pub const MEMTABLE_BULK_MERGE_THRESHOLD: &str = "memtable.bulk.merge_threshold";
64/// Option key for bulk memtable encode row threshold.
65pub const MEMTABLE_BULK_ENCODE_ROW_THRESHOLD: &str = "memtable.bulk.encode_row_threshold";
66/// Option key for bulk memtable encode bytes threshold.
67pub const MEMTABLE_BULK_ENCODE_BYTES_THRESHOLD: &str = "memtable.bulk.encode_bytes_threshold";
68/// Option key for bulk memtable max merge groups.
69pub const MEMTABLE_BULK_MAX_MERGE_GROUPS: &str = "memtable.bulk.max_merge_groups";
70/// Option key for memtable partition tree index max keys per shard.
71pub const MEMTABLE_PARTITION_TREE_INDEX_MAX_KEYS_PER_SHARD: &str =
72    "memtable.partition_tree.index_max_keys_per_shard";
73/// Option key for memtable partition tree data freeze threshold.
74pub const MEMTABLE_PARTITION_TREE_DATA_FREEZE_THRESHOLD: &str =
75    "memtable.partition_tree.data_freeze_threshold";
76/// Option key for memtable partition tree fork dictionary bytes.
77pub const MEMTABLE_PARTITION_TREE_FORK_DICTIONARY_BYTES: &str =
78    "memtable.partition_tree.fork_dictionary_bytes";
79/// Option key for skipping WAL.
80pub const SKIP_WAL_KEY: &str = "skip_wal";
81/// Option key for sst format.
82pub const SST_FORMAT_KEY: &str = "sst_format";
83/// Option key for the max number of rows in a parquet row group.
84pub const MAX_ROW_GROUP_ROW_COUNT: &str = "max_row_group_row_count";
85/// Upper bound for [`MAX_ROW_GROUP_ROW_COUNT`].
86pub const MAX_ROW_GROUP_ROW_COUNT_LIMIT: usize = 10 * 1024 * 1024;
87/// Option key for preserving per-row sequence numbers through flush and compaction.
88pub const PRESERVE_ROW_SEQUENCE: &str = "preserve_row_sequence";
89/// Option key for experimental Parquet float field encoding.
90pub const EXPERIMENTAL_SST_FLOAT_FIELD_ENCODING: &str = "experimental_sst_float_field_encoding";
91
92/// Encoding policy for direct floating-point field columns in Parquet SSTs.
93#[derive(
94    Debug,
95    Default,
96    Clone,
97    Copy,
98    PartialEq,
99    Eq,
100    Hash,
101    serde::Serialize,
102    serde::Deserialize,
103    strum::EnumString,
104)]
105#[serde(rename_all = "snake_case")]
106#[strum(serialize_all = "snake_case")]
107pub enum FloatFieldEncoding {
108    /// The existing Parquet writer behavior.
109    #[default]
110    Default,
111    /// Parquet byte-stream-split encoding.
112    ByteStreamSplit,
113}
114// Note: Adding new options here should also check if this option should be removed in [metric_engine::engine::create::region_options_for_metadata_region].
115
116/// Conflicting values supplied through the legacy and canonical TWCS trigger options.
117#[derive(Debug, PartialEq, Eq)]
118pub struct TwcsTriggerOptionConflict {
119    /// Value supplied under [`TWCS_TRIGGER_FILE_NUM`].
120    pub legacy_value: String,
121    /// Value supplied under [`TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM`].
122    pub canonical_value: String,
123}
124
125/// Normalizes the active-window TWCS trigger option to its legacy storage key.
126pub fn normalize_twcs_trigger_options(
127    options: &mut HashMap<String, String>,
128) -> Result<(), TwcsTriggerOptionConflict> {
129    let Some(canonical_value) = options.get(TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM).cloned() else {
130        return Ok(());
131    };
132    if let Some(legacy_value) = options.get(TWCS_TRIGGER_FILE_NUM)
133        && legacy_value != &canonical_value
134    {
135        return Err(TwcsTriggerOptionConflict {
136            legacy_value: legacy_value.clone(),
137            canonical_value,
138        });
139    }
140
141    options.remove(TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM);
142    options.insert(TWCS_TRIGGER_FILE_NUM.to_string(), canonical_value);
143    Ok(())
144}
145
146/// Returns true if the `key` is a valid option key for the mito engine.
147pub fn is_mito_engine_option_key(key: &str) -> bool {
148    [
149        "ttl",
150        WRITE_BUFFER_SIZE_KEY,
151        AUTO_FLUSH_INTERVAL_KEY,
152        COMPACTION_TYPE,
153        COMPACTION_OVERRIDE,
154        TWCS_TRIGGER_FILE_NUM,
155        TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM,
156        TWCS_ACTIVE_WINDOW_L1_MERGE_TRIGGER,
157        TWCS_INACTIVE_WINDOW_TRIGGER_FILE_NUM,
158        TWCS_INACTIVE_WINDOW_L1_MERGE_TRIGGER,
159        TWCS_MAX_OUTPUT_FILE_SIZE,
160        TWCS_TIME_WINDOW,
161        TWCS_REMOTE_COMPACTION,
162        TWCS_FALLBACK_TO_LOCAL,
163        "storage",
164        "index.inverted_index.ignore_column_ids",
165        "index.inverted_index.segment_row_count",
166        WAL_OPTIONS_KEY,
167        MEMTABLE_TYPE,
168        MEMTABLE_BULK_MERGE_THRESHOLD,
169        MEMTABLE_BULK_ENCODE_ROW_THRESHOLD,
170        MEMTABLE_BULK_ENCODE_BYTES_THRESHOLD,
171        MEMTABLE_BULK_MAX_MERGE_GROUPS,
172        MEMTABLE_PARTITION_TREE_INDEX_MAX_KEYS_PER_SHARD,
173        MEMTABLE_PARTITION_TREE_DATA_FREEZE_THRESHOLD,
174        MEMTABLE_PARTITION_TREE_FORK_DICTIONARY_BYTES,
175        // We don't allow to create a mito table with sparse primary key encoding directly.
176        APPEND_MODE_KEY,
177        MERGE_MODE_KEY,
178        SST_FORMAT_KEY,
179        MAX_ROW_GROUP_ROW_COUNT,
180        PRESERVE_ROW_SEQUENCE,
181        EXPERIMENTAL_SST_FLOAT_FIELD_ENCODING,
182    ]
183    .contains(&key)
184}
185
186#[cfg(test)]
187mod tests {
188    use std::collections::HashMap;
189
190    use super::*;
191
192    #[test]
193    fn test_is_mito_engine_option_key() {
194        assert!(is_mito_engine_option_key("ttl"));
195        assert!(is_mito_engine_option_key("write_buffer_size"));
196        assert!(is_mito_engine_option_key("auto_flush_interval"));
197        assert!(is_mito_engine_option_key("compaction.type"));
198        assert!(is_mito_engine_option_key("compaction.override"));
199        assert!(is_mito_engine_option_key(
200            "compaction.twcs.trigger_file_num"
201        ));
202        assert!(is_mito_engine_option_key(
203            "compaction.twcs.active_window.trigger_file_num"
204        ));
205        assert!(is_mito_engine_option_key(
206            "compaction.twcs.active_window.l1_merge_trigger"
207        ));
208        assert!(is_mito_engine_option_key(
209            "compaction.twcs.inactive_window.trigger_file_num"
210        ));
211        assert!(is_mito_engine_option_key(
212            "compaction.twcs.inactive_window.l1_merge_trigger"
213        ));
214        assert!(is_mito_engine_option_key("compaction.twcs.time_window"));
215        assert!(is_mito_engine_option_key("storage"));
216        assert!(is_mito_engine_option_key(
217            "index.inverted_index.ignore_column_ids"
218        ));
219        assert!(is_mito_engine_option_key(
220            "index.inverted_index.segment_row_count"
221        ));
222        assert!(is_mito_engine_option_key("wal_options"));
223        assert!(is_mito_engine_option_key("memtable.type"));
224        assert!(is_mito_engine_option_key("memtable.bulk.merge_threshold"));
225        assert!(is_mito_engine_option_key(
226            "memtable.bulk.encode_row_threshold"
227        ));
228        assert!(is_mito_engine_option_key(
229            "memtable.bulk.encode_bytes_threshold"
230        ));
231        assert!(is_mito_engine_option_key("memtable.bulk.max_merge_groups"));
232        assert!(is_mito_engine_option_key(
233            "memtable.partition_tree.index_max_keys_per_shard"
234        ));
235        assert!(is_mito_engine_option_key(
236            "memtable.partition_tree.data_freeze_threshold"
237        ));
238        assert!(is_mito_engine_option_key(
239            "memtable.partition_tree.fork_dictionary_bytes"
240        ));
241        assert!(is_mito_engine_option_key("append_mode"));
242        assert!(is_mito_engine_option_key("max_row_group_row_count"));
243        assert!(is_mito_engine_option_key("preserve_row_sequence"));
244        assert!(is_mito_engine_option_key(
245            EXPERIMENTAL_SST_FLOAT_FIELD_ENCODING
246        ));
247        assert!(!is_mito_engine_option_key("foo"));
248    }
249
250    #[test]
251    fn test_float_field_encoding_values() {
252        assert_eq!(
253            "default".parse::<FloatFieldEncoding>(),
254            Ok(FloatFieldEncoding::Default)
255        );
256        assert_eq!(
257            "byte_stream_split".parse::<FloatFieldEncoding>(),
258            Ok(FloatFieldEncoding::ByteStreamSplit)
259        );
260        assert_eq!(
261            serde_json::from_str::<FloatFieldEncoding>("\"byte_stream_split\"").unwrap(),
262            FloatFieldEncoding::ByteStreamSplit
263        );
264        assert!(serde_json::from_str::<FloatFieldEncoding>("\"unknown\"").is_err());
265    }
266
267    #[test]
268    fn test_normalize_twcs_trigger_aliases_to_legacy_key() {
269        let expected = HashMap::from([(TWCS_TRIGGER_FILE_NUM.to_string(), "4".to_string())]);
270        for mut options in [
271            HashMap::from([(TWCS_TRIGGER_FILE_NUM.to_string(), "4".to_string())]),
272            HashMap::from([(
273                TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM.to_string(),
274                "4".to_string(),
275            )]),
276            HashMap::from([
277                (TWCS_TRIGGER_FILE_NUM.to_string(), "4".to_string()),
278                (
279                    TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM.to_string(),
280                    "4".to_string(),
281                ),
282            ]),
283        ] {
284            normalize_twcs_trigger_options(&mut options).unwrap();
285            assert_eq!(expected, options);
286        }
287    }
288
289    #[test]
290    fn test_normalize_twcs_trigger_conflicting_aliases() {
291        let mut options = HashMap::from([
292            (TWCS_TRIGGER_FILE_NUM.to_string(), "4".to_string()),
293            (
294                TWCS_ACTIVE_WINDOW_TRIGGER_FILE_NUM.to_string(),
295                "8".to_string(),
296            ),
297        ]);
298        let original = options.clone();
299
300        let error = normalize_twcs_trigger_options(&mut options).unwrap_err();
301
302        assert_eq!("4", error.legacy_value);
303        assert_eq!("8", error.canonical_value);
304        assert_eq!(original, options);
305    }
306}
store_api/mito_engine_options.rs

store_api/
mito_engine_options.rs
