split the dumps between v1 and v2

This commit is contained in:
tamo 2021-04-27 10:27:43 +02:00
parent e389c088eb
commit c4d898a265
No known key found for this signature in database
GPG Key ID: 20CD8020AFA88D69
6 changed files with 205 additions and 66 deletions

View File

@ -55,10 +55,10 @@ impl ApiKeys {
}
impl Data {
pub fn new(options: Opt) -> anyhow::Result<Data> {
pub async fn new(options: Opt) -> anyhow::Result<Data> {
let path = options.db_path.clone();
let index_controller = IndexController::new(&path, &options)?;
let index_controller = IndexController::new(&path, &options).await?;
let mut api_keys = ApiKeys {
master: options.clone().master_key,

View File

@ -1,14 +1,13 @@
use std::{
fs::File,
path::{Path, PathBuf},
sync::Arc,
};
mod v1;
mod v2;
use std::{fs::File, path::{Path, PathBuf}, sync::Arc};
use anyhow::bail;
use heed::EnvOpenOptions;
use log::{error, info};
use milli::update::{IndexDocumentsMethod, UpdateBuilder, UpdateFormat};
use serde::{Deserialize, Serialize};
use serde::{de::Deserializer, Deserialize, Serialize};
use tempfile::TempDir;
use tokio::fs;
use tokio::task::spawn_blocking;
@ -20,13 +19,30 @@ use crate::index::Index;
use crate::index_controller::uuid_resolver;
use crate::{helpers::compression, index::Settings};
pub (super) fn deserialize_some<'de, T, D>(deserializer: D) -> Result<Option<T>, D::Error>
where
T: Deserialize<'de>,
D: Deserializer<'de>,
{
Deserialize::deserialize(deserializer).map(Some)
}
#[derive(Debug, Serialize, Deserialize, Copy, Clone)]
enum DumpVersion {
V1,
V2,
}
impl DumpVersion {
const CURRENT: Self = Self::V1;
const CURRENT: Self = Self::V2;
/// Select the good importation function from the `DumpVersion` of metadata
pub fn import_index(self, size: usize, dump_path: &Path, index_path: &Path) -> anyhow::Result<()> {
match self {
Self::V1 => v1::import_index(size, dump_path, index_path),
Self::V2 => v2::import_index(size, dump_path, index_path),
}
}
}
#[derive(Debug, Serialize, Deserialize)]
@ -141,16 +157,6 @@ where
}
}
/// Extract Settings from `settings.json` file present at provided `dir_path`
fn settings_from_path(dir_path: &Path) -> anyhow::Result<Settings> {
let path = dir_path.join("settings.json");
let file = File::open(path)?;
let reader = std::io::BufReader::new(file);
let metadata = serde_json::from_reader(reader)?;
Ok(metadata)
}
/// Write Settings in `settings.json` file at provided `dir_path`
fn settings_to_path(settings: &Settings, dir_path: &Path) -> anyhow::Result<()> {
let path = dir_path.join("settings.json");
@ -161,40 +167,7 @@ fn settings_to_path(settings: &Settings, dir_path: &Path) -> anyhow::Result<()>
Ok(())
}
fn import_index_v1(size: usize, dump_path: &Path, index_path: &Path) -> anyhow::Result<()> {
std::fs::create_dir_all(&index_path)?;
let mut options = EnvOpenOptions::new();
options.map_size(size);
let index = milli::Index::new(options, index_path)?;
let index = Index(Arc::new(index));
// extract `settings.json` file and import content
let settings = settings_from_path(&dump_path)?;
let update_builder = UpdateBuilder::new(0);
index.update_settings(&settings, update_builder)?;
let update_builder = UpdateBuilder::new(1);
let file = File::open(&index_path.join("documents.jsonl"))?;
let reader = std::io::BufReader::new(file);
index.update_documents(
UpdateFormat::JsonStream,
IndexDocumentsMethod::ReplaceDocuments,
reader,
update_builder,
None,
)?;
// the last step: we extract the milli::Index and close it
Arc::try_unwrap(index.0)
.map_err(|_e| "[dumps] At this point no one is supposed to have a reference on the index")
.unwrap()
.prepare_for_closing()
.wait();
Ok(())
}
pub fn load_dump(
pub async fn load_dump(
db_path: impl AsRef<Path>,
dump_path: impl AsRef<Path>,
size: usize,
@ -212,15 +185,10 @@ pub fn load_dump(
// read dump metadata
let metadata = DumpMetadata::from_path(&tmp_dir_path)?;
// choose importation function from DumpVersion of metadata
let import_index = match metadata.dump_version {
DumpVersion::V1 => import_index_v1,
};
// remove indexes which have same `uuid` than indexes to import and create empty indexes
let existing_index_uids = futures::executor::block_on(uuid_resolver.list())?;
let existing_index_uids = uuid_resolver.list().await?;
info!("Deleting indexes provided in the dump...");
info!("Deleting indexes already present in the db and provided in the dump...");
for idx in &metadata.indexes {
if let Some((_, uuid)) = existing_index_uids.iter().find(|(s, _)| s == &idx.uid) {
// if we find the index in the `uuid_resolver` it's supposed to exist on the file system
@ -237,18 +205,19 @@ pub fn load_dump(
}
} else {
// if the index does not exist in the `uuid_resolver` we create it
futures::executor::block_on(uuid_resolver.create(idx.uid.clone()))?;
uuid_resolver.create(idx.uid.clone()).await?;
}
}
// import each indexes content
for idx in metadata.indexes {
let dump_path = tmp_dir_path.join(&idx.uid);
let uuid = futures::executor::block_on(uuid_resolver.get(idx.uid))?;
let uuid = uuid_resolver.get(idx.uid).await?;
let index_path = db_path.join(&format!("indexes/index-{}", uuid));
let update_path = db_path.join(&format!("updates/updates-{}", uuid)); // TODO: add the update db
info!("Importing dump from {} into {}...", dump_path.display(), index_path.display());
import_index(size, &dump_path, &index_path).unwrap();
metadata.dump_version.import_index(size, &dump_path, &index_path).unwrap();
info!("Dump importation from {} succeed", dump_path.display());
}

View File

@ -0,0 +1,119 @@
use std::collections::{BTreeMap, BTreeSet};
use log::warn;
use serde::{Deserialize, Serialize};
use crate::index_controller;
use super::*;
/// This is the settings used in the last version of meilisearch exporting dump in V1
#[derive(Default, Clone, Serialize, Deserialize, Debug)]
#[serde(rename_all = "camelCase", deny_unknown_fields)]
struct Settings {
#[serde(default, deserialize_with = "deserialize_some")]
pub ranking_rules: Option<Option<Vec<String>>>,
#[serde(default, deserialize_with = "deserialize_some")]
pub distinct_attribute: Option<Option<String>>,
#[serde(default, deserialize_with = "deserialize_some")]
pub searchable_attributes: Option<Option<Vec<String>>>,
#[serde(default, deserialize_with = "deserialize_some")]
pub displayed_attributes: Option<Option<BTreeSet<String>>>,
#[serde(default, deserialize_with = "deserialize_some")]
pub stop_words: Option<Option<BTreeSet<String>>>,
#[serde(default, deserialize_with = "deserialize_some")]
pub synonyms: Option<Option<BTreeMap<String, Vec<String>>>>,
#[serde(default, deserialize_with = "deserialize_some")]
pub attributes_for_faceting: Option<Option<Vec<String>>>,
}
/// we need to **always** be able to convert the old settings to the settings currently being used
impl From<Settings> for index_controller::Settings {
fn from(settings: Settings) -> Self {
if settings.distinct_attribute.flatten().is_some() {
error!("`distinct_attribute` are not yet implemented and thus will be ignored");
}
if settings.synonyms.flatten().is_some() {
error!("`synonyms` are not yet implemented and thus will be ignored");
}
Self {
// we need to convert the old `Vec<String>` into a `BTreeSet<String>`
displayed_attributes: settings.displayed_attributes.map(|o| o.map(|vec| vec.into_iter().collect())),
searchable_attributes: settings.searchable_attributes,
// we previously had a `Vec<String>` but now we have a `HashMap<String, String>`
// representing the name of the faceted field + the type of the field. Since the type
// was not known in the V1 of the dump we are just going to assume everything is a
// String
attributes_for_faceting: settings.attributes_for_faceting.map(|o| o.map(|vec| vec.into_iter().map(|key| (key, String::from("string"))).collect())),
// we need to convert the old `Vec<String>` into a `BTreeSet<String>`
ranking_rules: settings.ranking_rules.map(|o| o.map(|vec| vec.into_iter().filter_map(|criterion| {
match criterion.as_str() {
"words" | "typo" | "proximity" => Some(criterion),
s if s.starts_with("asc") || s.starts_with("desc") => Some(criterion),
"wordsPosition" => {
warn!("The criteria `words` and `wordsPosition` have been merged into a single criterion `words` so `wordsPositon` will be ignored");
Some(String::from("words"))
}
"attribute" | "exactness" => {
error!("The criterion `{}` is not implemented currently and thus will be ignored", criterion);
None
}
s => {
error!("Unknown criterion found in the dump: `{}`, it will be ignored", s);
None
}
}
}).collect())),
// we need to convert the old `Vec<String>` into a `BTreeSet<String>`
stop_words: settings.stop_words.map(|o| o.map(|vec| vec.into_iter().collect())),
}
}
}
/// Extract Settings from `settings.json` file present at provided `dir_path`
fn import_settings(dir_path: &Path) -> anyhow::Result<Settings> {
let path = dir_path.join("settings.json");
let file = File::open(path)?;
let reader = std::io::BufReader::new(file);
let metadata = serde_json::from_reader(reader)?;
Ok(metadata)
}
pub fn import_index(size: usize, dump_path: &Path, index_path: &Path) -> anyhow::Result<()> {
info!("Importing a dump from an old version of meilisearch with dump version 1");
std::fs::create_dir_all(&index_path)?;
let mut options = EnvOpenOptions::new();
options.map_size(size);
let index = milli::Index::new(options, index_path)?;
let index = Index(Arc::new(index));
// extract `settings.json` file and import content
let settings = import_settings(&dump_path)?;
dbg!(&settings);
let settings = settings.into();
dbg!(&settings);
let update_builder = UpdateBuilder::new(0);
index.update_settings(&settings, update_builder)?;
let update_builder = UpdateBuilder::new(1);
let file = File::open(&dump_path.join("documents.jsonl"))?;
let reader = std::io::BufReader::new(file);
index.update_documents(
UpdateFormat::JsonStream,
IndexDocumentsMethod::ReplaceDocuments,
reader,
update_builder,
None,
)?;
// the last step: we extract the original milli::Index and close it
Arc::try_unwrap(index.0)
.map_err(|_e| "[dumps] At this point no one is supposed to have a reference on the index")
.unwrap()
.prepare_for_closing()
.wait();
Ok(())
}

View File

@ -0,0 +1,51 @@
use heed::EnvOpenOptions;
use milli::update::{IndexDocumentsMethod, UpdateBuilder, UpdateFormat};
use crate::index::Index;
use crate::index_controller::Settings;
use std::{fs::File, path::Path, sync::Arc};
/// Extract Settings from `settings.json` file present at provided `dir_path`
fn import_settings(dir_path: &Path) -> anyhow::Result<Settings> {
let path = dir_path.join("settings.json");
let file = File::open(path)?;
let reader = std::io::BufReader::new(file);
let metadata = serde_json::from_reader(reader)?;
Ok(metadata)
}
pub fn import_index(size: usize, dump_path: &Path, index_path: &Path) -> anyhow::Result<()> {
std::fs::create_dir_all(&index_path)?;
let mut options = EnvOpenOptions::new();
options.map_size(size);
let index = milli::Index::new(options, index_path)?;
let index = Index(Arc::new(index));
// extract `settings.json` file and import content
let settings = import_settings(&dump_path)?;
let update_builder = UpdateBuilder::new(0);
index.update_settings(&settings, update_builder)?;
dbg!(settings);
let update_builder = UpdateBuilder::new(1);
let file = File::open(&dump_path.join("documents.jsonl"))?;
let reader = std::io::BufReader::new(file);
index.update_documents(
UpdateFormat::JsonStream,
IndexDocumentsMethod::ReplaceDocuments,
reader,
update_builder,
None,
)?;
// the last step: we extract the original milli::Index and close it
Arc::try_unwrap(index.0)
.map_err(|_e| "[dumps] At this point no one is supposed to have a reference on the index")
.unwrap()
.prepare_for_closing()
.wait();
Ok(())
}

View File

@ -75,7 +75,7 @@ pub struct Stats {
}
impl IndexController {
pub fn new(path: impl AsRef<Path>, options: &Opt) -> anyhow::Result<Self> {
pub async fn new(path: impl AsRef<Path>, options: &Opt) -> anyhow::Result<Self> {
let index_size = options.max_mdb_size.get_bytes() as usize;
let update_store_size = options.max_udb_size.get_bytes() as usize;
@ -92,7 +92,7 @@ impl IndexController {
&options.db_path,
path,
index_size,
);
).await?;
}

View File

@ -54,7 +54,7 @@ async fn main() -> Result<(), MainError> {
//snapshot::load_snapshot(&opt.db_path, path, opt.ignore_snapshot_if_db_exists, opt.ignore_missing_snapshot)?;
//}
let data = Data::new(opt.clone())?;
let data = Data::new(opt.clone()).await?;
//if !opt.no_analytics {
//let analytics_data = data.clone();