Compare commits

...

51 Commits

Author SHA1 Message Date
ecd36b15f0 exposes all the s3 arguments 2023-09-13 18:17:56 +02:00
8a2e8a887f Load the latest snapshot when we start the engine 2023-09-12 18:08:24 +02:00
309c33a418 Fix again the dots 2023-09-12 17:55:01 +02:00
9b01506cee Move the load snapshot step into a function 2023-09-12 16:05:02 +02:00
f37fdceb15 Use slashes instead of dots for the s3 paths separators 2023-09-12 15:46:15 +02:00
f544cfa444 Remove tasks and content file on the s3 2023-09-12 15:19:45 +02:00
c158d03337 Fix internal error 2023-09-12 14:46:13 +02:00
b7109c0fd2 start a script to run everything 2023-09-12 11:34:59 +02:00
a53a0fdb77 Store content files into the S3 2023-09-11 18:17:22 +02:00
719fdd701b Fix and crash when the tasks path is unknown 2023-09-07 11:31:18 +02:00
01c13c98ac Mastering minio 2023-09-06 17:54:21 +02:00
5b89276fcc starts using s3 2023-09-05 19:25:09 +02:00
41697c4d65 Introduce the zk-tasks folder 2023-09-04 18:24:34 +02:00
7d85753573 Make the snapshot download work 2023-09-04 17:38:56 +02:00
76657af1f9 Add the options into the IndexScheduler 2023-09-04 16:38:05 +02:00
966cbdab69 make the tests compile again 2023-09-04 15:39:54 +02:00
0c68b9ed4c WIP making the final snapshot swap 2023-08-31 15:56:42 +02:00
d7233ecdb8 Make things to compile again 2023-08-31 14:55:14 +02:00
95a011af13 Wrap the IndexScheduler fields into an inner struct 2023-08-31 10:36:33 +02:00
e257710961 WIP fix the tests 2023-08-30 18:03:24 +02:00
9dd4423054 Fix the watcher ordering of the auth/ node 2023-08-30 17:51:22 +02:00
8c3ad57ef9 React to changes towards the cluster members 2023-08-30 17:40:12 +02:00
2d1434da81 Keep the ZK flow when enqueuing tasks 2023-08-30 17:15:15 +02:00
c488a4a351 Fixup a lot of small issues on the ZK config 2023-08-30 16:42:55 +02:00
0c7d7c68bc WIP moving to the sync zookeeper API 2023-08-30 15:06:12 +02:00
854745c670 wip: starts working on importing the snapshots 2023-08-16 18:41:05 +02:00
777eebb759 starts creating snapshot, the import is still missing 2023-08-10 15:00:25 +02:00
61ccfaf9bc wake up after registering a task 2023-08-10 09:39:39 +02:00
f0c4d36ff7 implement the deletion of tasks after processing a batch
add a lot of comments and logs
2023-08-10 09:36:43 +02:00
8c20d6e2fe fix the leader election 2023-08-09 17:23:13 +02:00
8e437ed76c Start leader election and task processing (WIP) 2023-08-09 16:52:38 +02:00
1191ec5939 fix the register task watcher 2023-08-08 13:18:55 +02:00
0d20d08daf fix a few warnings 2023-08-08 11:39:48 +02:00
b66bf049b5 Create a task on zookeeper side when a task is created locally 2023-08-07 17:02:51 +02:00
b2f36b9b97 Comment Meilisearch container by default 2023-08-07 17:02:00 +02:00
b311089435 Update zookeeper client 2023-08-07 14:20:01 +02:00
3d46e84d97 update docker compose as an example 2023-08-03 17:15:24 +02:00
ad7f8edff8 fix auto-synchronization with zk 2023-08-03 14:43:29 +02:00
5ce01bcb53 add logs 2023-08-03 13:59:05 +02:00
d5523cc6ac fix the tests 2023-08-03 12:28:08 +02:00
fe7a312ec6 Import the already existing api keys on startup 2023-08-03 12:25:32 +02:00
57dc4b148c implement the watcher for all kind of operations 2023-08-03 10:52:13 +02:00
a325ddfe6a Forward the key deletions to zookeeper 2023-08-03 10:36:49 +02:00
0cd81573b4 Forward the keys update to zookeeper 2023-08-03 10:22:34 +02:00
b0ff595f60 Event Listener: delete local key if deleted on ZK 2023-08-02 18:36:36 +02:00
3eb6f4b56f Create api keys 2023-08-02 16:52:45 +02:00
49f976c8d8 fix analitics compilation 2023-08-02 14:17:03 +02:00
84d56f3320 send the creation of api-key to zookeeper 2023-08-02 13:57:30 +02:00
97e3dfd99d makes zk available inside the auth-controller with config coming from the cli, it compiles 2023-08-02 13:17:40 +02:00
dc38da95c4 WIP 2023-08-02 12:00:02 +02:00
2ce8b42757 REMOVE: add docker compĂ´se for tests 2023-08-02 11:59:54 +02:00
30 changed files with 2038 additions and 843 deletions

433
Cargo.lock generated
View File

@ -9,7 +9,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "617a8268e3537fe1d8c9ead925fca49ef6400927ee7bc26750e90ecee14ce4b8"
dependencies = [
"bitflags 1.3.2",
"bytes",
"bytes 1.4.0",
"futures-core",
"futures-sink",
"memchr",
@ -49,7 +49,7 @@ dependencies = [
"base64 0.21.2",
"bitflags 1.3.2",
"brotli",
"bytes",
"bytes 1.4.0",
"bytestring",
"derive_more",
"encoding_rs",
@ -119,7 +119,7 @@ dependencies = [
"actix-utils",
"futures-core",
"futures-util",
"mio",
"mio 0.8.7",
"num_cpus",
"socket2",
"tokio",
@ -182,9 +182,9 @@ dependencies = [
"actix-utils",
"actix-web-codegen",
"ahash 0.7.6",
"bytes",
"bytes 1.4.0",
"bytestring",
"cfg-if",
"cfg-if 1.0.0",
"cookie",
"derive_more",
"encoding_rs",
@ -251,7 +251,7 @@ version = "0.8.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "433cfd6710c9986c576a25ca913c39d66a6474107b406f34f91d4a8923395241"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"cipher",
"cpufeatures",
]
@ -273,7 +273,7 @@ version = "0.8.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2c99f64d1e06488f620f932677e24bc6e2897582980441ae90a671415bd7ec2f"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"getrandom",
"once_cell",
"version_check",
@ -428,12 +428,54 @@ dependencies = [
"critical-section",
]
[[package]]
name = "attohttpc"
version = "0.22.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1fcf00bc6d5abb29b5f97e3c61a90b6d3caa12f3faf897d4a3e3607c050a35a7"
dependencies = [
"http",
"log",
"rustls 0.20.8",
"serde",
"serde_json",
"url",
"webpki",
"webpki-roots 0.22.6",
]
[[package]]
name = "autocfg"
version = "1.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d468802bab17cbc0cc575e9b053f41e72aa36bfa6b7f55e3529ffa43161b97fa"
[[package]]
name = "aws-creds"
version = "0.34.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3776743bb68d4ad02ba30ba8f64373f1be4e082fe47651767171ce75bb2f6cf5"
dependencies = [
"attohttpc",
"dirs",
"log",
"quick-xml",
"rust-ini",
"serde",
"thiserror",
"time",
"url",
]
[[package]]
name = "aws-region"
version = "0.25.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "056557a61427d0e5ba29dd931031c8ffed4ee7a550e7cd55692a9d8deb0a9dba"
dependencies = [
"thiserror",
]
[[package]]
name = "backtrace"
version = "0.3.67"
@ -442,7 +484,7 @@ checksum = "233d376d6d185f2a3093e58f283f60f880315b6c60075b01f36b3b85154564ca"
dependencies = [
"addr2line",
"cc",
"cfg-if",
"cfg-if 1.0.0",
"libc",
"miniz_oxide 0.6.2",
"object",
@ -472,7 +514,7 @@ name = "benchmarks"
version = "1.3.0"
dependencies = [
"anyhow",
"bytes",
"bytes 1.4.0",
"convert_case 0.6.0",
"criterion",
"csv",
@ -612,6 +654,12 @@ version = "1.4.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "14c189c53d098945499cdfa7ecc63567cf3886b3332b312a5b4585d8d3a6a610"
[[package]]
name = "bytes"
version = "0.5.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0e4cec68f03f32e44924783795810fa50a7035d8c8ebe78580ad7e6c703fba38"
[[package]]
name = "bytes"
version = "1.4.0"
@ -624,7 +672,7 @@ version = "1.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "238e4886760d98c4f899360c834fa93e62cf7f721ac3c2da375cbdf4b8679aae"
dependencies = [
"bytes",
"bytes 1.4.0",
]
[[package]]
@ -682,6 +730,12 @@ dependencies = [
"smallvec",
]
[[package]]
name = "cfg-if"
version = "0.1.10"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "4785bdd1c96b2a846b2bd7cc02e86b6b3dbf14e7e53446c4f54c92a361040822"
[[package]]
name = "cfg-if"
version = "1.0.0"
@ -897,7 +951,7 @@ version = "1.3.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b540bd8bc810d3885c6ea91e2018302f68baba2129ab3e88f32389ee9370880d"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
]
[[package]]
@ -948,7 +1002,7 @@ version = "0.8.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2801af0d36612ae591caa9568261fddce32ce6e08a7275ea334a06a4ad021a2c"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"crossbeam-channel",
"crossbeam-deque",
"crossbeam-epoch",
@ -962,7 +1016,7 @@ version = "0.5.8"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a33c2bf77f2df06183c3aa30d1e96c0695a313d4f9c453cc3762a6db39f99200"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"crossbeam-utils",
]
@ -972,7 +1026,7 @@ version = "0.8.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ce6fd6f855243022dcecf8702fef0c297d4338e226845fe067f6341ad9fa0cef"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"crossbeam-epoch",
"crossbeam-utils",
]
@ -984,7 +1038,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "46bd5f3f85273295a9d14aedfb86f6aadbff6d8f5295c4a9edb08e819dcf5695"
dependencies = [
"autocfg",
"cfg-if",
"cfg-if 1.0.0",
"crossbeam-utils",
"memoffset",
"scopeguard",
@ -996,7 +1050,7 @@ version = "0.3.8"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d1cfb3ea8a53f37c40dea2c7bedcbd88bdfae54f5e2175d6ecaff1c988353add"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"crossbeam-utils",
]
@ -1006,7 +1060,7 @@ version = "0.8.15"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3c063cd8cc95f5c377ed0d4b49a4b21f632396ff690e8470c29b3359b346984b"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
]
[[package]]
@ -1176,16 +1230,36 @@ dependencies = [
"subtle",
]
[[package]]
name = "dirs"
version = "4.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ca3aa72a6f96ea37bbc5aa912f6788242832f75369bdfdadcb0e38423f100059"
dependencies = [
"dirs-sys",
]
[[package]]
name = "dirs-next"
version = "1.0.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "cf36e65a80337bea855cd4ef9b8401ffce06a7baedf2e85ec467b1ac3f6e82b6"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"dirs-sys-next",
]
[[package]]
name = "dirs-sys"
version = "0.3.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1b1d1d91c932ef41c0f2663aa8b0ca0342d444d842c06914aa0a7e352d0bada6"
dependencies = [
"libc",
"redox_users",
"winapi 0.3.9",
]
[[package]]
name = "dirs-sys-next"
version = "0.1.2"
@ -1194,9 +1268,15 @@ checksum = "4ebda144c4fe02d1f7ea1a7d9641b6fc6b580adcfa024ae48797ecdeb6825b4d"
dependencies = [
"libc",
"redox_users",
"winapi",
"winapi 0.3.9",
]
[[package]]
name = "dlv-list"
version = "0.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0688c2a7f92e427f44895cd63841bff7b29f8d7a1648b9e7e07a4a365b2e1257"
[[package]]
name = "dump"
version = "1.3.0"
@ -1219,14 +1299,14 @@ dependencies = [
"tempfile",
"thiserror",
"time",
"uuid 1.3.3",
"uuid 1.4.1",
]
[[package]]
name = "either"
version = "1.8.1"
version = "1.9.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7fcaabb2fef8c910e7f4c7ce9f67a1283a1715879a7c230ca9d6d1ae31f16d91"
checksum = "a26ae43d7bcc3b814de94796a5e736d4029efb0ee900c12e2d54c993ad1a1e07"
dependencies = [
"serde",
]
@ -1307,7 +1387,7 @@ version = "0.8.32"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "071a31f4ee85403370b58aca746f01041ede6f0da2730960ad001edc2b71b394"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
]
[[package]]
@ -1418,7 +1498,7 @@ dependencies = [
"faux",
"tempfile",
"thiserror",
"uuid 1.3.3",
"uuid 1.4.1",
]
[[package]]
@ -1427,7 +1507,7 @@ version = "0.2.21"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5cbc844cecaee9d4443931972e1289c8ff485cb4cc2767cb03ca139ed6885153"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"libc",
"redox_syscall 0.2.16",
"windows-sys 0.48.0",
@ -1481,6 +1561,22 @@ version = "0.4.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7ab85b9b05e3978cc9a9cf8fea7f01b494e1a09ed3037e16ba39edc7a29eb61a"
[[package]]
name = "fuchsia-zircon"
version = "0.3.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2e9763c69ebaae630ba35f74888db465e49e259ba1bc0eda7d06f4a067615d82"
dependencies = [
"bitflags 1.3.2",
"fuchsia-zircon-sys",
]
[[package]]
name = "fuchsia-zircon-sys"
version = "0.3.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3dcaa9ae7725d12cdb85b3ad99a434db70b468c09ded17e012d86b5c1010f7a7"
[[package]]
name = "futures"
version = "0.3.28"
@ -1614,7 +1710,7 @@ version = "0.2.9"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c85e1d9ab2eadba7e5040d4e09cbd6d072b76a557ad64e797c2cb9d4da21d7e4"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"libc",
"wasi",
]
@ -1673,7 +1769,7 @@ version = "0.3.19"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d357c7ae988e7d2182f7d7871d0b963962420b0678b0997ce7de72001aeab782"
dependencies = [
"bytes",
"bytes 1.4.0",
"fnv",
"futures-core",
"futures-sink",
@ -1805,7 +1901,7 @@ version = "0.2.9"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "bd6effc99afb63425aff9b05836f029929e345a6148a14b7ecd5ab67af944482"
dependencies = [
"bytes",
"bytes 1.4.0",
"fnv",
"itoa",
]
@ -1816,7 +1912,7 @@ version = "0.4.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d5f38f16d184e36f2408a55281cd658ecbd3ca05cce6d6510a176eca393e26d1"
dependencies = [
"bytes",
"bytes 1.4.0",
"http",
"pin-project-lite",
]
@ -1845,7 +1941,7 @@ version = "0.14.26"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ab302d72a6f11a3b910431ff93aae7e773078c769f0a3ef15fb9ec692ed147d4"
dependencies = [
"bytes",
"bytes 1.4.0",
"futures-channel",
"futures-core",
"futures-util",
@ -1912,15 +2008,19 @@ dependencies = [
"meilisearch-types",
"nelson",
"page_size 0.5.0",
"parking_lot",
"puffin",
"roaring",
"rust-s3",
"serde",
"serde_json",
"synchronoise",
"tempfile",
"thiserror",
"time",
"uuid 1.3.3",
"tokio",
"uuid 1.4.1",
"zookeeper",
]
[[package]]
@ -1975,7 +2075,7 @@ version = "0.1.12"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7a5bbe824c507c5da5956355e86a746d82e0e1464f65d862cc5e71da70e94b2c"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
]
[[package]]
@ -2004,6 +2104,15 @@ dependencies = [
"windows-sys 0.48.0",
]
[[package]]
name = "iovec"
version = "0.1.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b2b3ea6ff95e175473f8ffe6a7eb7c00d054240321b84c57051175fe3c1e075e"
dependencies = [
"libc",
]
[[package]]
name = "ipnet"
version = "2.7.2"
@ -2102,6 +2211,16 @@ dependencies = [
"simple_asn1",
]
[[package]]
name = "kernel32-sys"
version = "0.2.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7507624b29483431c0ba2d82aece8ca6cdba9382bff4ddd0f7490560c056098d"
dependencies = [
"winapi 0.2.8",
"winapi-build",
]
[[package]]
name = "language-tags"
version = "0.3.2"
@ -2114,6 +2233,12 @@ version = "1.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e2abad23fbc42b3700f2f279844dc832adb2b2eb069b2df918f455c4e18cc646"
[[package]]
name = "lazycell"
version = "1.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "830d08ce1d1d941e6b30645f1a0eb5643013d835ce3779a5fc208261dbe10f55"
[[package]]
name = "levenshtein_automata"
version = "0.2.1"
@ -2492,6 +2617,17 @@ version = "1.0.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3e2e65a1a2e43cfcb47a895c4c8b10d1f4a61097f9f254f183aee60cad9c651d"
[[package]]
name = "maybe-async"
version = "0.2.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0f1b8c13cb1f814b634a96b2c725449fe7ed464a7b8781de8688be5ffbd3f305"
dependencies = [
"proc-macro2",
"quote",
"syn 1.0.109",
]
[[package]]
name = "md5"
version = "0.7.0"
@ -2524,7 +2660,7 @@ dependencies = [
"brotli",
"bstr",
"byte-unit",
"bytes",
"bytes 1.4.0",
"cargo_toml",
"clap",
"crossbeam-channel",
@ -2569,6 +2705,7 @@ dependencies = [
"rayon",
"regex",
"reqwest",
"rust-s3",
"rustls 0.20.8",
"rustls-pemfile",
"segment",
@ -2591,11 +2728,12 @@ dependencies = [
"tokio-stream",
"toml",
"urlencoding",
"uuid 1.3.3",
"uuid 1.4.1",
"vergen",
"walkdir",
"yaup",
"zip",
"zookeeper",
]
[[package]]
@ -2605,6 +2743,7 @@ dependencies = [
"base64 0.21.2",
"enum-iterator",
"hmac",
"log",
"maplit",
"meilisearch-types",
"rand",
@ -2614,7 +2753,8 @@ dependencies = [
"sha2",
"thiserror",
"time",
"uuid 1.3.3",
"uuid 1.4.1",
"zookeeper",
]
[[package]]
@ -2644,7 +2784,7 @@ dependencies = [
"thiserror",
"time",
"tokio",
"uuid 1.3.3",
"uuid 1.4.1",
]
[[package]]
@ -2724,7 +2864,7 @@ dependencies = [
"tempfile",
"thiserror",
"time",
"uuid 1.3.3",
"uuid 1.4.1",
]
[[package]]
@ -2776,6 +2916,25 @@ dependencies = [
"adler",
]
[[package]]
name = "mio"
version = "0.6.23"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "4afd66f5b91bf2a3bc13fad0e21caedac168ca4c707504e75585648ae80e4cc4"
dependencies = [
"cfg-if 0.1.10",
"fuchsia-zircon",
"fuchsia-zircon-sys",
"iovec",
"kernel32-sys",
"libc",
"log",
"miow",
"net2",
"slab",
"winapi 0.2.8",
]
[[package]]
name = "mio"
version = "0.8.7"
@ -2788,11 +2947,46 @@ dependencies = [
"windows-sys 0.48.0",
]
[[package]]
name = "mio-extras"
version = "2.0.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "52403fe290012ce777c4626790c8951324a2b9e3316b3143779c72b029742f19"
dependencies = [
"lazycell",
"log",
"mio 0.6.23",
"slab",
]
[[package]]
name = "miow"
version = "0.2.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ebd808424166322d4a38da87083bfddd3ac4c131334ed55856112eb06d46944d"
dependencies = [
"kernel32-sys",
"net2",
"winapi 0.2.8",
"ws2_32-sys",
]
[[package]]
name = "nelson"
version = "0.1.0"
source = "git+https://github.com/meilisearch/nelson.git?rev=675f13885548fb415ead8fbb447e9e6d9314000a#675f13885548fb415ead8fbb447e9e6d9314000a"
[[package]]
name = "net2"
version = "0.2.39"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b13b648036a2339d06de780866fbdfda0dde886de7b3af2ddeba8b14f4ee34ac"
dependencies = [
"cfg-if 0.1.10",
"libc",
"winapi 0.3.9",
]
[[package]]
name = "nom"
version = "7.1.3"
@ -2820,7 +3014,7 @@ version = "0.4.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e8a3895c6391c39d7fe7ebc444a87eb2991b2a0bc718fdabd071eec617fc68e4"
dependencies = [
"winapi",
"winapi 0.3.9",
]
[[package]]
@ -2900,6 +3094,16 @@ dependencies = [
"num-traits",
]
[[package]]
name = "ordered-multimap"
version = "0.4.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ccd746e37177e1711c20dd619a1620f34f5c8b569c53590a72dedd5344d8924a"
dependencies = [
"dlv-list",
"hashbrown 0.12.3",
]
[[package]]
name = "page_size"
version = "0.4.2"
@ -2907,7 +3111,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "eebde548fbbf1ea81a99b128872779c437752fb99f217c45245e1a61dcd9edcd"
dependencies = [
"libc",
"winapi",
"winapi 0.3.9",
]
[[package]]
@ -2917,7 +3121,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1b7663cbd190cfd818d08efa8497f6cd383076688c49a391ef7c0d03cd12b561"
dependencies = [
"libc",
"winapi",
"winapi 0.3.9",
]
[[package]]
@ -2936,7 +3140,7 @@ version = "0.9.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9069cbb9f99e3a5083476ccb29ceb1de18b9118cafa53e90c9551235de2b9521"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"libc",
"redox_syscall 0.2.16",
"smallvec",
@ -3211,7 +3415,7 @@ version = "0.13.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "449811d15fbdf5ceb5c1144416066429cf82316e2ec8ce0c1f6f8a02e7bbcf8c"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"fnv",
"lazy_static",
"libc",
@ -3237,7 +3441,7 @@ dependencies = [
"anyhow",
"bincode",
"byteorder",
"cfg-if",
"cfg-if 1.0.0",
"instant",
"lz4_flex",
"once_cell",
@ -3257,6 +3461,16 @@ dependencies = [
"puffin",
]
[[package]]
name = "quick-xml"
version = "0.26.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7f50b1c63b38611e7d4d7f68b82d3ad0cc71a2ad2e7f61fc10f1328d917c93cd"
dependencies = [
"memchr",
"serde",
]
[[package]]
name = "quote"
version = "1.0.30"
@ -3387,7 +3601,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "cde824a14b7c14f85caff81225f411faacc04a2013f41670f41443742b1c1c55"
dependencies = [
"base64 0.21.2",
"bytes",
"bytes 1.4.0",
"encoding_rs",
"futures-core",
"futures-util",
@ -3437,7 +3651,7 @@ dependencies = [
"spin 0.5.2",
"untrusted",
"web-sys",
"winapi",
"winapi 0.3.9",
]
[[package]]
@ -3464,6 +3678,45 @@ dependencies = [
"smallvec",
]
[[package]]
name = "rust-ini"
version = "0.18.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f6d5f2436026b4f6e79dc829837d467cc7e9a55ee40e750d716713540715a2df"
dependencies = [
"cfg-if 1.0.0",
"ordered-multimap",
]
[[package]]
name = "rust-s3"
version = "0.33.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1b2ac5ff6acfbe74226fa701b5ef793aaa054055c13ebb7060ad36942956e027"
dependencies = [
"async-trait",
"attohttpc",
"aws-creds",
"aws-region",
"base64 0.13.1",
"bytes 1.4.0",
"cfg-if 1.0.0",
"hex",
"hmac",
"http",
"log",
"maybe-async",
"md5",
"percent-encoding",
"quick-xml",
"serde",
"serde_derive",
"sha2",
"thiserror",
"time",
"url",
]
[[package]]
name = "rustc-demangle"
version = "0.1.23"
@ -3697,7 +3950,7 @@ version = "0.10.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f5058ada175748e33390e40e872bd0fe59a19f265d0158daa551c5a88a76009c"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"cpufeatures",
"digest",
]
@ -3708,7 +3961,7 @@ version = "0.10.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f04293dc80c3993519f2d7f6f511707ee7094fe0c6d3406feb330cdb3540eba3"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"cpufeatures",
"digest",
]
@ -3719,7 +3972,7 @@ version = "0.10.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "82e6b795fe2e3b1e845bafcb27aa35405c4d47cdfc92af5fc8d3002f76cebdc0"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"cpufeatures",
"digest",
]
@ -3799,6 +4052,12 @@ dependencies = [
"version_check",
]
[[package]]
name = "snowflake"
version = "1.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "27207bb65232eda1f588cf46db2fee75c0808d557f6b3cf19a75f5d6d7c94df1"
[[package]]
name = "socket2"
version = "0.4.9"
@ -3806,7 +4065,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "64a4a911eed85daf18834cfaa86a79b7d266ff93ff5ba14005426219480ed662"
dependencies = [
"libc",
"winapi",
"winapi 0.3.9",
]
[[package]]
@ -3908,13 +4167,13 @@ version = "0.28.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b4c2f3ca6693feb29a89724516f016488e9aafc7f37264f898593ee4b942f31b"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"core-foundation-sys",
"libc",
"ntapi",
"once_cell",
"rayon",
"winapi",
"winapi 0.3.9",
]
[[package]]
@ -3943,7 +4202,7 @@ version = "3.5.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b9fbec84f381d5795b08656e4912bec604d162bff9291d6189a78f4c8ab87998"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"fastrand",
"redox_syscall 0.3.5",
"rustix 0.37.19",
@ -4038,9 +4297,9 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "94d7b1cfd2aa4011f2de74c2c4c63665e27a71006b0a192dcd2710272e73dfa2"
dependencies = [
"autocfg",
"bytes",
"bytes 1.4.0",
"libc",
"mio",
"mio 0.8.7",
"num_cpus",
"parking_lot",
"pin-project-lite",
@ -4099,7 +4358,7 @@ version = "0.7.8"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "806fe8c2c87eccc8b3267cbae29ed3ab2d0bd37fca70ab622e46aaa9375ddb7d"
dependencies = [
"bytes",
"bytes 1.4.0",
"futures-core",
"futures-sink",
"pin-project-lite",
@ -4153,7 +4412,7 @@ version = "0.1.37"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8ce8c33a8d48bd45d624a6e523445fd21ec13d3653cd51f681abf67418f54eb8"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"log",
"pin-project-lite",
"tracing-core",
@ -4289,9 +4548,9 @@ dependencies = [
[[package]]
name = "uuid"
version = "1.3.3"
version = "1.4.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "345444e32442451b267fc254ae85a209c64be56d2890e601a0c37ff0c3c5ecd2"
checksum = "79daa5ed5740825c40b389c5e50312b9c86df53fccd33f281df655642b43869d"
dependencies = [
"getrandom",
"serde",
@ -4310,7 +4569,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f21b881cd6636ece9735721cf03c1fe1e774fe258683d084bb2812ab67435749"
dependencies = [
"anyhow",
"cfg-if",
"cfg-if 1.0.0",
"enum-iterator",
"getset",
"git2",
@ -4368,7 +4627,7 @@ version = "0.2.86"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5bba0e8cb82ba49ff4e229459ff22a191bbe9a1cb3a341610c9c33efc27ddf73"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"wasm-bindgen-macro",
]
@ -4393,7 +4652,7 @@ version = "0.4.36"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2d1985d03709c53167ce907ff394f5316aa22cb4e12761295c5dc57dacb6297e"
dependencies = [
"cfg-if",
"cfg-if 1.0.0",
"js-sys",
"wasm-bindgen",
"web-sys",
@ -4476,6 +4735,12 @@ dependencies = [
"once_cell",
]
[[package]]
name = "winapi"
version = "0.2.8"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "167dc9d6949a9b857f3451275e911c3f44255842c1f7a76f33c55103a909087a"
[[package]]
name = "winapi"
version = "0.3.9"
@ -4486,6 +4751,12 @@ dependencies = [
"winapi-x86_64-pc-windows-gnu",
]
[[package]]
name = "winapi-build"
version = "0.1.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2d315eee3b34aca4797b2da6b13ed88266e6d612562a0c46390af8299fc699bc"
[[package]]
name = "winapi-i686-pc-windows-gnu"
version = "0.4.0"
@ -4498,7 +4769,7 @@ version = "0.1.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "70ec6ce85bb158151cae5e5c87f95a8e97d2c0c4b001223f33a334e3ce5de178"
dependencies = [
"winapi",
"winapi 0.3.9",
]
[[package]]
@ -4654,7 +4925,17 @@ version = "0.10.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "80d0f4e272c85def139476380b12f9ac60926689dd2e01d4923222f40580869d"
dependencies = [
"winapi",
"winapi 0.3.9",
]
[[package]]
name = "ws2_32-sys"
version = "0.2.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d59cefebd0c892fa2dd6de581e937301d8552cb44489cdff035c6187cb63fa5e"
dependencies = [
"winapi 0.2.8",
"winapi-build",
]
[[package]]
@ -4732,6 +5013,32 @@ dependencies = [
"zstd 0.11.2+zstd.1.5.2",
]
[[package]]
name = "zookeeper"
version = "0.8.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2312b424380193701a7341cec0551b80d2e3afd827ea0d3440af67899156ce10"
dependencies = [
"byteorder",
"bytes 0.5.6",
"lazy_static",
"log",
"mio 0.6.23",
"mio-extras",
"snowflake",
"zookeeper_derive",
]
[[package]]
name = "zookeeper_derive"
version = "0.4.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "42307291e3c8b2e4082e5647572da863f0470511d0ecb1618a4cd0a361549723"
dependencies = [
"quote",
"syn 1.0.109",
]
[[package]]
name = "zstd"
version = "0.11.2+zstd.1.5.2"

View File

@ -36,24 +36,3 @@ opt-level = 3
opt-level = 3
[profile.dev.package.roaring]
opt-level = 3
[profile.dev.package.lindera-ipadic-builder]
opt-level = 3
[profile.dev.package.encoding]
opt-level = 3
[profile.dev.package.yada]
opt-level = 3
[profile.release.package.lindera-ipadic-builder]
opt-level = 3
[profile.release.package.encoding]
opt-level = 3
[profile.release.package.yada]
opt-level = 3
[profile.bench.package.lindera-ipadic-builder]
opt-level = 3
[profile.bench.package.encoding]
opt-level = 3
[profile.bench.package.yada]
opt-level = 3

59
docker-compose.yml Normal file
View File

@ -0,0 +1,59 @@
version: "3.9"
services:
zk1:
container_name: zk1
hostname: zk1
image: bitnami/zookeeper:3.7.1
ports:
- 21811:2181
environment:
- ALLOW_ANONYMOUS_LOGIN=yes
- ZOO_SERVER_ID=1
- ZOO_SERVERS=0.0.0.0:2888:3888,zk2:2888:3888,zk3:2888:3888
zk2:
container_name: zk2
hostname: zk2
image: bitnami/zookeeper:3.7.1
ports:
- 21812:2181
environment:
- ALLOW_ANONYMOUS_LOGIN=yes
- ZOO_SERVER_ID=2
- ZOO_SERVERS=zk1:2888:3888,0.0.0.0:2888:3888,zk3:2888:3888
zk3:
container_name: zk3
hostname: zk3
image: bitnami/zookeeper:3.7.1
ports:
- 21813:2181
environment:
- ALLOW_ANONYMOUS_LOGIN=yes
- ZOO_SERVER_ID=3
- ZOO_SERVERS=zk1:2888:3888,zk2:2888:3888,0.0.0.0:2888:3888
zoonavigator:
container_name: zoonavigator
image: elkozmon/zoonavigator
ports:
- 9000:9000
# Meilisearch instances
# m1:
# container_name: m1
# hostname: m1
# image: getmeili/meilisearch:prototype-zookeeper-ha-0
# ports:
# - 7700:7700
# environment:
# - MEILI_ZK_URL=zk1:2181
# - MEILI_MASTER_KEY=masterkey
# restart: always
# m2:
# container_name: m2
# hostname: m2
# image: getmeili/meilisearch:prototype-zookeeper-ha-0
# ports:
# - 7701:7700
# environment:
# - MEILI_ZK_URL=zk2:2181
# - MEILI_MASTER_KEY=masterkey
# restart: always

View File

@ -22,20 +22,6 @@ pub enum Error {
pub type Result<T> = std::result::Result<T, Error>;
impl Deref for File {
type Target = NamedTempFile;
fn deref(&self) -> &Self::Target {
&self.file
}
}
impl DerefMut for File {
fn deref_mut(&mut self) -> &mut Self::Target {
&mut self.file
}
}
#[derive(Clone, Debug)]
pub struct FileStore {
path: PathBuf,
@ -146,6 +132,20 @@ impl File {
}
}
impl Deref for File {
type Target = NamedTempFile;
fn deref(&self) -> &Self::Target {
&self.file
}
}
impl DerefMut for File {
fn deref_mut(&mut self) -> &mut Self::Target {
&mut self.file
}
}
#[cfg(test)]
mod test {
use std::io::Write;

61
ha_test/run.sh Normal file
View File

@ -0,0 +1,61 @@
#!/bin/bash
function is_everything_installed {
everything_ok=yes
if hash zkli 2>/dev/null; then
echo "âś… zkli installed"
else
everything_ok=no
echo "🥺 zkli is missing, please run \`cargo install zkli\`"
fi
if hash s3cmd 2>/dev/null; then
echo "âś… s3cmd installed"
else
everything_ok=no
echo "🥺 s3cmd is missing, see how to install it here https://s3tools.org/s3cmd"
fi
if [ $everything_ok = "no" ]; then
echo "Exiting..."
exit 1
fi
}
# param: addr of zookeeper
function connect_to_zookeeper {
if ! zkli -a "$1" ls > /dev/null; then
echo "zkli can't connect"
return 1
fi
}
# param: addr of the s3 bucket
function connect_to_s3 {
# S3_SECRET_KEY
# S3_ACCESS_KEY
# S3_HOST
# S3_BUCKET
s3cmd --host="$S3_HOST" --host-bucket="$S3_BUCKET" --access_key="$ACCESS_KEY" --secret_key="$S3_SECRET_KEY" ls
if $?; then
echo "s3cmd can't connect"
return 1
fi
}
is_everything_installed
ZOOKEEPER_ADDR="localhost:2181"
if ! connect_to_zookeeper $ZOOKEEPER_ADDR; then
ZOOKEEPER_ADDR="localhost:21811"
if ! connect_to_zookeeper $ZOOKEEPER_ADDR; then
echo "Can't connect to zkli"
exit 1
fi
fi
connect_to_s3

View File

@ -31,6 +31,10 @@ tempfile = "3.5.0"
thiserror = "1.0.40"
time = { version = "0.3.20", features = ["serde-well-known", "formatting", "parsing", "macros"] }
uuid = { version = "1.3.1", features = ["serde", "v4"] }
tokio = { version = "1.27.0", features = ["full"] }
zookeeper = "0.8.0"
parking_lot = "0.12.1"
rust-s3 = { version = "0.33.0", default-features = false, features = ["sync-rustls-tls"] }
[dev-dependencies]
big_s = "1.0.2"

View File

@ -43,7 +43,7 @@ use uuid::Uuid;
use crate::autobatcher::{self, BatchKind};
use crate::utils::{self, swap_index_uid_in_task};
use crate::{Error, IndexScheduler, ProcessingTasks, Result, TaskId};
use crate::{Error, IndexSchedulerInner, ProcessingTasks, Result, TaskId};
/// Represents a combination of tasks that can all be processed at the same time.
///
@ -198,6 +198,35 @@ impl Batch {
| IndexDocumentDeletionByFilter { index_uid, .. } => Some(index_uid),
}
}
/// Return the content fields uuids associated with this batch.
pub fn content_uuids(&self) -> Vec<Uuid> {
match self {
Batch::TaskCancelation { .. }
| Batch::TaskDeletion(_)
| Batch::Dump(_)
| Batch::IndexCreation { .. }
| Batch::IndexDocumentDeletionByFilter { .. }
| Batch::IndexUpdate { .. }
| Batch::SnapshotCreation(_)
| Batch::IndexDeletion { .. }
| Batch::IndexSwap { .. } => vec![],
Batch::IndexOperation { op, .. } => match op {
IndexOperation::DocumentOperation { operations, .. } => operations
.iter()
.flat_map(|op| match op {
DocumentOperation::Add(uuid) => Some(*uuid),
DocumentOperation::Delete(_) => None,
})
.collect(),
IndexOperation::DocumentDeletion { .. }
| IndexOperation::Settings { .. }
| IndexOperation::DocumentClear { .. }
| IndexOperation::SettingsAndDocumentOperation { .. }
| IndexOperation::DocumentClearAndSetting { .. } => vec![],
},
}
}
}
impl IndexOperation {
@ -213,7 +242,7 @@ impl IndexOperation {
}
}
impl IndexScheduler {
impl IndexSchedulerInner {
/// Convert an [`BatchKind`](crate::autobatcher::BatchKind) into a [`Batch`].
///
/// ## Arguments
@ -480,8 +509,7 @@ impl IndexScheduler {
if let Some(task_id) = to_cancel.max() {
// We retrieve the tasks that were processing before this tasks cancelation started.
// We must *not* reset the processing tasks before calling this method.
let ProcessingTasks { started_at, processing } =
&*self.processing_tasks.read().unwrap();
let ProcessingTasks { started_at, processing, .. } = &*self.processing_tasks.read();
return Ok(Some(Batch::TaskCancelation {
task: self.get_task(rtxn, task_id)?.ok_or(Error::CorruptedTaskQueue)?,
previous_started_at: *started_at,
@ -1392,7 +1420,7 @@ impl IndexScheduler {
fn delete_matched_tasks(&self, wtxn: &mut RwTxn, matched_tasks: &RoaringBitmap) -> Result<u64> {
// 1. Remove from this list the tasks that we are not allowed to delete
let enqueued_tasks = self.get_status(wtxn, Status::Enqueued)?;
let processing_tasks = &self.processing_tasks.read().unwrap().processing.clone();
let processing_tasks = &self.processing_tasks.read().processing.clone();
let all_task_ids = self.all_task_ids(wtxn)?;
let mut to_delete_tasks = all_task_ids & matched_tasks;

View File

@ -295,6 +295,11 @@ impl IndexMap {
"Attempt to finish deletion of an index that was being closed"
);
}
/// Returns the indexes that were opened by the `IndexMap`.
pub fn clear(&mut self) -> Vec<Index> {
self.available.clear().into_iter().map(|(_, (_, index))| index).collect()
}
}
/// Create or open an index in the specified path.
@ -335,7 +340,8 @@ mod tests {
impl IndexMapper {
fn test() -> (Self, Env, IndexSchedulerHandle) {
let (index_scheduler, handle) = IndexScheduler::test(true, vec![]);
(index_scheduler.index_mapper, index_scheduler.env, handle)
let index_scheduler = index_scheduler.inner();
(index_scheduler.index_mapper.clone(), index_scheduler.env.clone(), handle)
}
}

View File

@ -61,7 +61,7 @@ pub struct IndexMapper {
pub(crate) index_stats: Database<UuidCodec, SerdeJson<IndexStats>>,
/// Path to the folder where the LMDB environments of each index are.
base_path: PathBuf,
pub(crate) base_path: PathBuf,
/// The map size an index is opened with on the first time.
index_base_map_size: usize,
/// The quantity by which the map size of an index is incremented upon reopening, in bytes.
@ -135,7 +135,7 @@ impl IndexMapper {
index_growth_amount: usize,
index_count: usize,
enable_mdb_writemap: bool,
indexer_config: IndexerConfig,
indexer_config: Arc<IndexerConfig>,
) -> Result<Self> {
let mut wtxn = env.write_txn()?;
let index_mapping = env.create_database(&mut wtxn, Some(INDEX_MAPPING))?;
@ -150,7 +150,7 @@ impl IndexMapper {
index_base_map_size,
index_growth_amount,
enable_mdb_writemap,
indexer_config: Arc::new(indexer_config),
indexer_config,
})
}
@ -428,6 +428,11 @@ impl IndexMapper {
Ok(())
}
/// Returns the indexes that were opened by the `IndexMapper`.
pub fn clear(&mut self) -> Vec<Index> {
self.index_map.write().unwrap().clear()
}
/// The stats of an index.
///
/// If available in the cache, they are directly returned.

View File

@ -1,5 +1,6 @@
use std::collections::BTreeSet;
use std::fmt::Write;
use std::ops::Deref;
use meilisearch_types::heed::types::{OwnedType, SerdeBincode, SerdeJson, Str};
use meilisearch_types::heed::{Database, RoTxn};
@ -8,12 +9,13 @@ use meilisearch_types::tasks::{Details, Task};
use roaring::RoaringBitmap;
use crate::index_mapper::IndexMapper;
use crate::{IndexScheduler, Kind, Status, BEI128};
use crate::{IndexScheduler, IndexSchedulerInner, Kind, Status, BEI128};
pub fn snapshot_index_scheduler(scheduler: &IndexScheduler) -> String {
scheduler.assert_internally_consistent();
let IndexScheduler {
let inner = scheduler.inner();
let IndexSchedulerInner {
autobatching_enabled,
must_stop_processing: _,
processing_tasks,
@ -38,13 +40,15 @@ pub fn snapshot_index_scheduler(scheduler: &IndexScheduler) -> String {
test_breakpoint_sdr: _,
planned_failures: _,
run_loop_iteration: _,
} = scheduler;
zookeeper: _,
options: _,
} = inner.deref();
let rtxn = env.read_txn().unwrap();
let mut snap = String::new();
let processing_tasks = processing_tasks.read().unwrap().processing.clone();
let processing_tasks = processing_tasks.read().processing.clone();
snap.push_str(&format!("### Autobatching Enabled = {autobatching_enabled}\n"));
snap.push_str("### Processing Tasks:\n");
snap.push_str(&snapshot_bitmap(&processing_tasks));

File diff suppressed because it is too large Load Diff

View File

@ -1,5 +1,6 @@
//! Thread-safe `Vec`-backend LRU cache using [`std::sync::atomic::AtomicU64`] for synchronization.
use std::mem;
use std::sync::atomic::{AtomicU64, Ordering};
/// Thread-safe `Vec`-backend LRU cache
@ -190,6 +191,11 @@ where
}
None
}
/// Returns the generation associated to the key and values of the `LruMap`.
pub fn clear(&mut self) -> Vec<(AtomicU64, (K, V))> {
mem::take(&mut self.0.data)
}
}
/// The result of an insertion in a LRU map.

View File

@ -10,9 +10,9 @@ use meilisearch_types::tasks::{Details, IndexSwap, Kind, KindWithContent, Status
use roaring::{MultiOps, RoaringBitmap};
use time::OffsetDateTime;
use crate::{Error, IndexScheduler, Result, Task, TaskId, BEI128};
use crate::{Error, IndexSchedulerInner, Result, Task, TaskId, BEI128};
impl IndexScheduler {
impl IndexSchedulerInner {
pub(crate) fn all_task_ids(&self, rtxn: &RoTxn) -> Result<RoaringBitmap> {
enum_iterator::all().map(|s| self.get_status(rtxn, s)).union()
}
@ -331,11 +331,12 @@ pub fn clamp_to_page_size(size: usize) -> usize {
}
#[cfg(test)]
impl IndexScheduler {
impl crate::IndexScheduler {
/// Asserts that the index scheduler's content is internally consistent.
pub fn assert_internally_consistent(&self) {
let rtxn = self.env.read_txn().unwrap();
for task in self.all_tasks.iter(&rtxn).unwrap() {
let this = self.inner();
let rtxn = this.env.read_txn().unwrap();
for task in this.all_tasks.iter(&rtxn).unwrap() {
let (task_id, task) = task.unwrap();
let task_id = task_id.get();
@ -354,21 +355,21 @@ impl IndexScheduler {
} = task;
assert_eq!(uid, task.uid);
if let Some(task_index_uid) = &task_index_uid {
assert!(self
assert!(this
.index_tasks
.get(&rtxn, task_index_uid.as_str())
.unwrap()
.unwrap()
.contains(task.uid));
}
let db_enqueued_at = self
let db_enqueued_at = this
.enqueued_at
.get(&rtxn, &BEI128::new(enqueued_at.unix_timestamp_nanos()))
.unwrap()
.unwrap();
assert!(db_enqueued_at.contains(task_id));
if let Some(started_at) = started_at {
let db_started_at = self
let db_started_at = this
.started_at
.get(&rtxn, &BEI128::new(started_at.unix_timestamp_nanos()))
.unwrap()
@ -376,7 +377,7 @@ impl IndexScheduler {
assert!(db_started_at.contains(task_id));
}
if let Some(finished_at) = finished_at {
let db_finished_at = self
let db_finished_at = this
.finished_at
.get(&rtxn, &BEI128::new(finished_at.unix_timestamp_nanos()))
.unwrap()
@ -384,9 +385,9 @@ impl IndexScheduler {
assert!(db_finished_at.contains(task_id));
}
if let Some(canceled_by) = canceled_by {
let db_canceled_tasks = self.get_status(&rtxn, Status::Canceled).unwrap();
let db_canceled_tasks = this.get_status(&rtxn, Status::Canceled).unwrap();
assert!(db_canceled_tasks.contains(uid));
let db_canceling_task = self.get_task(&rtxn, canceled_by).unwrap().unwrap();
let db_canceling_task = this.get_task(&rtxn, canceled_by).unwrap().unwrap();
assert_eq!(db_canceling_task.status, Status::Succeeded);
match db_canceling_task.kind {
KindWithContent::TaskCancelation { query: _, tasks } => {
@ -427,7 +428,7 @@ impl IndexScheduler {
Details::IndexInfo { primary_key: pk1 } => match &kind {
KindWithContent::IndexCreation { index_uid, primary_key: pk2 }
| KindWithContent::IndexUpdate { index_uid, primary_key: pk2 } => {
self.index_tasks
this.index_tasks
.get(&rtxn, index_uid.as_str())
.unwrap()
.unwrap()
@ -535,23 +536,23 @@ impl IndexScheduler {
}
}
assert!(self.get_status(&rtxn, status).unwrap().contains(uid));
assert!(self.get_kind(&rtxn, kind.as_kind()).unwrap().contains(uid));
assert!(this.get_status(&rtxn, status).unwrap().contains(uid));
assert!(this.get_kind(&rtxn, kind.as_kind()).unwrap().contains(uid));
if let KindWithContent::DocumentAdditionOrUpdate { content_file, .. } = kind {
match status {
Status::Enqueued | Status::Processing => {
assert!(self
assert!(this
.file_store
.all_uuids()
.unwrap()
.any(|uuid| uuid.as_ref().unwrap() == &content_file),
"Could not find uuid `{content_file}` in the file_store. Available uuids are {:?}.",
self.file_store.all_uuids().unwrap().collect::<std::result::Result<Vec<_>, file_store::Error>>().unwrap(),
this.file_store.all_uuids().unwrap().collect::<std::result::Result<Vec<_>, file_store::Error>>().unwrap(),
);
}
Status::Succeeded | Status::Failed | Status::Canceled => {
assert!(self
assert!(this
.file_store
.all_uuids()
.unwrap()

View File

@ -14,6 +14,7 @@ license.workspace = true
base64 = "0.21.0"
enum-iterator = "1.4.0"
hmac = "0.12.1"
log = "0.4.19"
maplit = "1.0.2"
meilisearch-types = { path = "../meilisearch-types" }
rand = "0.8.5"
@ -24,3 +25,4 @@ sha2 = "0.10.6"
thiserror = "1.0.40"
time = { version = "0.3.20", features = ["serde-well-known", "formatting", "parsing", "macros"] }
uuid = { version = "1.3.1", features = ["serde", "v4"] }
zookeeper = "0.8.0"

View File

@ -19,6 +19,7 @@ internal_error!(
AuthControllerError: meilisearch_types::milli::heed::Error,
std::io::Error,
serde_json::Error,
zookeeper::ZkError,
std::str::Utf8Error
);

View File

@ -16,22 +16,113 @@ pub use store::open_auth_store_env;
use store::{generate_key_as_hexa, HeedAuthStore};
use time::OffsetDateTime;
use uuid::Uuid;
use zookeeper::{
Acl, AddWatchMode, CreateMode, WatchedEvent, WatchedEventType, ZkError, ZooKeeper,
};
#[derive(Clone)]
pub struct AuthController {
store: Arc<HeedAuthStore>,
master_key: Option<String>,
zookeeper: Option<Arc<ZooKeeper>>,
}
impl AuthController {
pub fn new(db_path: impl AsRef<Path>, master_key: &Option<String>) -> Result<Self> {
pub fn new(
db_path: impl AsRef<Path>,
master_key: &Option<String>,
zookeeper: Option<Arc<ZooKeeper>>,
) -> Result<Self> {
let store = HeedAuthStore::new(db_path)?;
let controller = Self { store: Arc::new(store), master_key: master_key.clone(), zookeeper };
if store.is_empty()? {
generate_default_keys(&store)?;
match controller.zookeeper {
// setup the auth zk environment, the `auth` node
Some(ref zookeeper) => {
// Zookeeper Event listener loop
let controller_clone = controller.clone();
let zkk = zookeeper.clone();
zookeeper.add_watch("/auth", AddWatchMode::PersistentRecursive, move |event| {
let WatchedEvent { event_type, path, keeper_state: _ } = dbg!(event);
match event_type {
WatchedEventType::NodeDeleted => {
// TODO: ugly unwraps
let path = path.unwrap();
let uuid = path.strip_prefix("/auth/").unwrap();
let uuid = Uuid::parse_str(&uuid).unwrap();
log::info!("The key {} has been deleted", uuid);
controller_clone.store.delete_api_key(uuid).unwrap();
}
WatchedEventType::NodeCreated | WatchedEventType::NodeDataChanged => {
let path = path.unwrap();
if path.strip_prefix("/auth/").map_or(false, |s| !s.is_empty()) {
let (key, _stat) = zkk.get_data(&path, false).unwrap();
let key: Key = serde_json::from_slice(&key).unwrap();
log::info!("The key {} has been deleted", key.uid);
controller_clone.store.put_api_key(key).unwrap();
}
}
otherwise => panic!("Got the unexpected `{otherwise:?}` event!"),
}
})?;
// TODO: we should catch the potential unexpected errors here https://docs.rs/zookeeper-client/latest/zookeeper_client/struct.Client.html#method.create
// for the moment we consider that `create` only returns Error::NodeExists.
match zookeeper.create(
"/auth",
vec![],
Acl::open_unsafe().clone(),
CreateMode::Persistent,
) {
// If the store is empty, we must generate and push the default api-keys.
Ok(_) => generate_default_keys(&controller)?,
// If the node exist we should clear our DB and download all the existing api-keys
Err(ZkError::NodeExists) => {
log::warn!("Auth directory already exists, we need to clear our keys + import the one in zookeeper");
let store = controller.store.clone();
store.delete_all_keys()?;
let children = zookeeper
.get_children("/auth", false)
.expect("Internal, the auth directory was deleted during execution.");
log::info!("Importing {} api-keys", children.len());
for path in children {
log::info!(" Importing {}", path);
match zookeeper.get_data(&format!("/auth/{}", &path), false) {
Ok((key, _stat)) => {
let key = serde_json::from_slice(&key).unwrap();
let store = controller.store.clone();
store.put_api_key(key)?;
}
Err(e) => panic!("{e}"),
}
// else the file was deleted while we were inserting the key. We ignore it.
// TODO: What happens if someone updates the files before we have the time
// to setup the watcher
}
}
e @ Err(
ZkError::NoNode | ZkError::NoChildrenForEphemerals | ZkError::InvalidACL,
) => unreachable!("{e:?}"),
Err(e) => panic!("{e}"),
}
// TODO: Race condition above:
// What happens if two node join exactly at the same moment:
// One will create the dir
// The second one will delete its DB, load nothing and install a watcher
// The first one will push its keys and should wake up and update the second one.
// / BUT, if the second one delete its DB and the first one push its files before the second one install the watcher we're fucked
}
None => {
if controller.store.is_empty()? {
generate_default_keys(&controller)?;
}
}
}
Ok(Self { store: Arc::new(store), master_key: master_key.clone() })
Ok(controller)
}
/// Return `Ok(())` if the auth controller is able to access one of its database.
@ -53,7 +144,24 @@ impl AuthController {
pub fn create_key(&self, create_key: CreateApiKey) -> Result<Key> {
match self.store.get_api_key(create_key.uid)? {
Some(_) => Err(AuthControllerError::ApiKeyAlreadyExists(create_key.uid.to_string())),
None => self.store.put_api_key(create_key.to_key()),
None => self.put_key(create_key.to_key()),
}
}
pub fn put_key(&self, key: Key) -> Result<Key> {
let store = self.store.clone();
match &self.zookeeper {
Some(zookeeper) => {
zookeeper.create(
&format!("/auth/{}", key.uid),
serde_json::to_vec_pretty(&key)?,
Acl::open_unsafe().clone(),
CreateMode::Persistent,
)?;
Ok(key)
}
None => store.put_api_key(key),
}
}
@ -68,7 +176,20 @@ impl AuthController {
name => key.name = name.set(),
};
key.updated_at = OffsetDateTime::now_utc();
self.store.put_api_key(key)
let store = self.store.clone();
// TODO: we may commit only after zk persisted the keys
match &self.zookeeper {
Some(zookeeper) => {
zookeeper.set_data(
&format!("/auth/{}", key.uid),
serde_json::to_vec_pretty(&key)?,
None,
)?;
Ok(key)
}
None => store.put_api_key(key),
}
}
pub fn get_key(&self, uid: Uuid) -> Result<Key> {
@ -110,7 +231,19 @@ impl AuthController {
}
pub fn delete_key(&self, uid: Uuid) -> Result<()> {
if self.store.delete_api_key(uid)? {
let deleted = match &self.zookeeper {
Some(zookeeper) => match zookeeper.delete(&format!("/auth/{}", uid), None) {
Ok(()) => true,
Err(ZkError::NoNode) => false,
Err(e) => return Err(e.into()),
},
None => {
let store = self.store.clone();
store.delete_api_key(uid)?
}
};
if deleted {
Ok(())
} else {
Err(AuthControllerError::ApiKeyNotFound(uid.to_string()))
@ -159,7 +292,7 @@ impl AuthController {
self.store.delete_all_keys()
}
/// Delete all the keys in the DB.
/// Insert a key in the DB without any check on its validity
pub fn raw_insert_key(&mut self, key: Key) -> Result<()> {
self.store.put_api_key(key)?;
Ok(())
@ -304,10 +437,9 @@ pub struct IndexSearchRules {
pub filter: Option<serde_json::Value>,
}
fn generate_default_keys(store: &HeedAuthStore) -> Result<()> {
store.put_api_key(Key::default_admin())?;
store.put_api_key(Key::default_search())?;
fn generate_default_keys(controller: &AuthController) -> Result<()> {
controller.put_key(Key::default_admin())?;
controller.put_key(Key::default_search())?;
Ok(())
}

View File

@ -80,6 +80,7 @@ reqwest = { version = "0.11.16", features = [
], default-features = false }
rustls = "0.20.8"
rustls-pemfile = "1.0.2"
rust-s3 = { version = "0.33.0", default-features = false, features = ["sync-rustls-tls"] }
segment = { version = "0.2.2", optional = true }
serde = { version = "1.0.160", features = ["derive"] }
serde_json = { version = "1.0.95", features = ["preserve_order"] }
@ -105,6 +106,7 @@ walkdir = "2.3.3"
yaup = "0.2.1"
serde_urlencoded = "0.7.1"
termcolor = "1.2.0"
zookeeper = "0.8.0"
[dev-dependencies]
actix-rt = "2.8.0"

View File

@ -312,6 +312,8 @@ impl From<Opt> for Infos {
config_file_path,
#[cfg(all(not(debug_assertions), feature = "analytics"))]
no_analytics: _,
zk_url: _,
s3_url: _,
} = options;
let schedule_snapshot = match schedule_snapshot {

View File

@ -39,6 +39,9 @@ use meilisearch_types::versioning::{check_version_file, create_version_file};
use meilisearch_types::{compression, milli, VERSION_FILE_NAME};
pub use option::Opt;
use option::ScheduleSnapshot;
use s3::creds::Credentials;
use s3::{Bucket, Region};
use zookeeper::ZooKeeper;
use crate::error::MeilisearchHttpError;
@ -136,14 +139,17 @@ enum OnFailure {
KeepDb,
}
pub fn setup_meilisearch(opt: &Opt) -> anyhow::Result<(Arc<IndexScheduler>, Arc<AuthController>)> {
pub async fn setup_meilisearch(
opt: &Opt,
zookeeper: Option<Arc<ZooKeeper>>,
) -> anyhow::Result<(Arc<IndexScheduler>, Arc<AuthController>)> {
let empty_db = is_empty_db(&opt.db_path);
let (index_scheduler, auth_controller) = if let Some(ref snapshot_path) = opt.import_snapshot {
let snapshot_path_exists = snapshot_path.exists();
// the db is empty and the snapshot exists, import it
if empty_db && snapshot_path_exists {
match compression::from_tar_gz(snapshot_path, &opt.db_path) {
Ok(()) => open_or_create_database_unchecked(opt, OnFailure::RemoveDb)?,
Ok(()) => open_or_create_database_unchecked(opt, OnFailure::RemoveDb, zookeeper)?,
Err(e) => {
std::fs::remove_dir_all(&opt.db_path)?;
return Err(e);
@ -160,14 +166,14 @@ pub fn setup_meilisearch(opt: &Opt) -> anyhow::Result<(Arc<IndexScheduler>, Arc<
bail!("snapshot doesn't exist at {}", snapshot_path.display())
// the snapshot and the db exist, and we can ignore the snapshot because of the ignore_snapshot_if_db_exists flag
} else {
open_or_create_database(opt, empty_db)?
open_or_create_database(opt, empty_db, zookeeper)?
}
} else if let Some(ref path) = opt.import_dump {
let src_path_exists = path.exists();
// the db is empty and the dump exists, import it
if empty_db && src_path_exists {
let (mut index_scheduler, mut auth_controller) =
open_or_create_database_unchecked(opt, OnFailure::RemoveDb)?;
open_or_create_database_unchecked(opt, OnFailure::RemoveDb, zookeeper)?;
match import_dump(&opt.db_path, path, &mut index_scheduler, &mut auth_controller) {
Ok(()) => (index_scheduler, auth_controller),
Err(e) => {
@ -187,10 +193,10 @@ pub fn setup_meilisearch(opt: &Opt) -> anyhow::Result<(Arc<IndexScheduler>, Arc<
// the dump and the db exist and we can ignore the dump because of the ignore_dump_if_db_exists flag
// or, the dump is missing but we can ignore that because of the ignore_missing_dump flag
} else {
open_or_create_database(opt, empty_db)?
open_or_create_database(opt, empty_db, zookeeper)?
}
} else {
open_or_create_database(opt, empty_db)?
open_or_create_database(opt, empty_db, zookeeper)?
};
// We create a loop in a thread that registers snapshotCreation tasks
@ -199,15 +205,12 @@ pub fn setup_meilisearch(opt: &Opt) -> anyhow::Result<(Arc<IndexScheduler>, Arc<
if let ScheduleSnapshot::Enabled(snapshot_delay) = opt.schedule_snapshot {
let snapshot_delay = Duration::from_secs(snapshot_delay);
let index_scheduler = index_scheduler.clone();
thread::Builder::new()
.name(String::from("register-snapshot-tasks"))
.spawn(move || loop {
thread::sleep(snapshot_delay);
if let Err(e) = index_scheduler.register(KindWithContent::SnapshotCreation) {
error!("Error while registering snapshot: {}", e);
}
})
.unwrap();
thread::spawn(move || loop {
thread::sleep(snapshot_delay);
if let Err(e) = index_scheduler.register(KindWithContent::SnapshotCreation) {
error!("Error while registering snapshot: {}", e);
}
});
}
Ok((index_scheduler, auth_controller))
@ -217,34 +220,52 @@ pub fn setup_meilisearch(opt: &Opt) -> anyhow::Result<(Arc<IndexScheduler>, Arc<
fn open_or_create_database_unchecked(
opt: &Opt,
on_failure: OnFailure,
zookeeper: Option<Arc<ZooKeeper>>,
) -> anyhow::Result<(IndexScheduler, AuthController)> {
// we don't want to create anything in the data.ms yet, thus we
// wrap our two builders in a closure that'll be executed later.
let auth_controller = AuthController::new(&opt.db_path, &opt.master_key);
let auth_controller = AuthController::new(&opt.db_path, &opt.master_key, zookeeper.clone());
let instance_features = opt.to_instance_features();
let index_scheduler_builder = || -> anyhow::Result<_> {
Ok(IndexScheduler::new(IndexSchedulerOptions {
version_file_path: opt.db_path.join(VERSION_FILE_NAME),
auth_path: opt.db_path.join("auth"),
tasks_path: opt.db_path.join("tasks"),
update_file_path: opt.db_path.join("update_files"),
indexes_path: opt.db_path.join("indexes"),
snapshots_path: opt.snapshot_dir.clone(),
dumps_path: opt.dump_dir.clone(),
task_db_size: opt.max_task_db_size.get_bytes() as usize,
index_base_map_size: opt.max_index_size.get_bytes() as usize,
enable_mdb_writemap: opt.experimental_reduce_indexing_memory_usage,
indexer_config: (&opt.indexer_options).try_into()?,
autobatching_enabled: true,
max_number_of_tasks: 1_000_000,
index_growth_amount: byte_unit::Byte::from_str("10GiB").unwrap().get_bytes() as usize,
index_count: DEFAULT_INDEX_COUNT,
instance_features,
})?)
};
let index_scheduler = IndexScheduler::new(Arc::new(IndexSchedulerOptions {
version_file_path: opt.db_path.join(VERSION_FILE_NAME),
auth_path: opt.db_path.join("auth"),
tasks_path: opt.db_path.join("tasks"),
update_file_path: opt.db_path.join("update_files"),
indexes_path: opt.db_path.join("indexes"),
snapshots_path: opt.snapshot_dir.clone(),
dumps_path: opt.dump_dir.clone(),
task_db_size: opt.max_task_db_size.get_bytes() as usize,
index_base_map_size: opt.max_index_size.get_bytes() as usize,
enable_mdb_writemap: opt.experimental_reduce_indexing_memory_usage,
indexer_config: (&opt.indexer_options).try_into().map(Arc::new)?,
autobatching_enabled: true,
max_number_of_tasks: 1_000_000,
index_growth_amount: byte_unit::Byte::from_str("10GiB").unwrap().get_bytes() as usize,
index_count: DEFAULT_INDEX_COUNT,
instance_features,
zookeeper: zookeeper.clone(),
s3: opt.s3_url.as_ref().map(|url| {
Arc::new(
Bucket::new(
opt.s3_bucket.as_deref().unwrap(),
Region::Custom { region: opt.s3_region.clone(), endpoint: url.clone() },
Credentials {
access_key: opt.s3_access_key.clone(),
secret_key: opt.s3_secret_key.clone(),
security_token: opt.s3_security_token.clone(),
session_token: None,
expiration: None,
},
)
.unwrap()
.with_path_style(),
)
}),
}))
.map_err(anyhow::Error::from);
match (
index_scheduler_builder(),
index_scheduler,
auth_controller.map_err(anyhow::Error::from),
create_version_file(&opt.db_path).map_err(anyhow::Error::from),
) {
@ -262,12 +283,13 @@ fn open_or_create_database_unchecked(
fn open_or_create_database(
opt: &Opt,
empty_db: bool,
zookeeper: Option<Arc<ZooKeeper>>,
) -> anyhow::Result<(IndexScheduler, AuthController)> {
if !empty_db {
check_version_file(&opt.db_path)?;
}
open_or_create_database_unchecked(opt, OnFailure::KeepDb)
open_or_create_database_unchecked(opt, OnFailure::KeepDb, zookeeper)
}
fn import_dump(
@ -277,6 +299,7 @@ fn import_dump(
auth: &mut AuthController,
) -> Result<(), anyhow::Error> {
let reader = File::open(dump_path)?;
let index_scheduler = index_scheduler.inner();
let mut dump_reader = dump::DumpReader::open(reader)?;
if let Some(date) = dump_reader.date() {

View File

@ -2,6 +2,7 @@ use std::env;
use std::io::{stderr, Write};
use std::path::PathBuf;
use std::sync::Arc;
use std::time::Duration;
use actix_web::http::KeepAlive;
use actix_web::web::Data;
@ -12,6 +13,7 @@ use meilisearch::analytics::Analytics;
use meilisearch::{analytics, create_app, prototype_name, setup_meilisearch, Opt};
use meilisearch_auth::{generate_master_key, AuthController, MASTER_KEY_MIN_SIZE};
use termcolor::{Color, ColorChoice, ColorSpec, StandardStream, WriteColor};
use zookeeper::ZooKeeper;
#[global_allocator]
static ALLOC: mimalloc::MiMalloc = mimalloc::MiMalloc;
@ -63,7 +65,10 @@ async fn main() -> anyhow::Result<()> {
_ => (),
}
let (index_scheduler, auth_controller) = setup_meilisearch(&opt)?;
let timeout = Duration::from_millis(2500);
let zookeeper =
opt.zk_url.as_ref().map(|url| Arc::new(ZooKeeper::connect(url, timeout, drop).unwrap()));
let (index_scheduler, auth_controller) = setup_meilisearch(&opt, zookeeper).await?;
#[cfg(all(not(debug_assertions), feature = "analytics"))]
let analytics = if !opt.no_analytics {

View File

@ -28,6 +28,13 @@ const MEILI_DB_PATH: &str = "MEILI_DB_PATH";
const MEILI_HTTP_ADDR: &str = "MEILI_HTTP_ADDR";
const MEILI_MASTER_KEY: &str = "MEILI_MASTER_KEY";
const MEILI_ENV: &str = "MEILI_ENV";
const MEILI_ZK_URL: &str = "MEILI_ZK_URL";
const MEILI_S3_URL: &str = "MEILI_S3_URL";
const MEILI_S3_BUCKET: &str = "MEILI_S3_BUCKET";
const MEILI_S3_ACCESS_KEY: &str = "MEILI_S3_ACCESS_KEY";
const MEILI_S3_SECRET_KEY: &str = "MEILI_S3_SECRET_KEY";
const MEILI_S3_SECURITY_TOKEN: &str = "MEILI_S3_SECURITY_TOKEN";
const MEILI_S3_REGION: &str = "MEILI_S3_REGION";
#[cfg(all(not(debug_assertions), feature = "analytics"))]
const MEILI_NO_ANALYTICS: &str = "MEILI_NO_ANALYTICS";
const MEILI_HTTP_PAYLOAD_SIZE_LIMIT: &str = "MEILI_HTTP_PAYLOAD_SIZE_LIMIT";
@ -56,6 +63,7 @@ const DEFAULT_CONFIG_FILE_PATH: &str = "./config.toml";
const DEFAULT_DB_PATH: &str = "./data.ms";
const DEFAULT_HTTP_ADDR: &str = "localhost:7700";
const DEFAULT_ENV: &str = "development";
const DEFAULT_S3_REGION: &str = "eu-central-1";
const DEFAULT_HTTP_PAYLOAD_SIZE_LIMIT: &str = "100 MB";
const DEFAULT_SNAPSHOT_DIR: &str = "snapshots/";
const DEFAULT_SNAPSHOT_INTERVAL_SEC: u64 = 86400;
@ -154,6 +162,36 @@ pub struct Opt {
#[serde(default = "default_env")]
pub env: String,
/// Sets the HTTP address and port used to communicate with the zookeeper cluster.
/// If ran locally, the default url is `http://localhost:2181/`.
#[clap(long, env = MEILI_ZK_URL)]
pub zk_url: Option<String>,
/// Sets the address and port used to communicate with the S3 bucket.
#[clap(long, env = MEILI_S3_URL)]
pub s3_url: Option<String>,
/// Sets the region used to communicate with the s3 bucket.
#[clap(long, env = MEILI_S3_REGION, default_value_t = default_s3_region())]
#[serde(default = "default_s3_region")]
pub s3_region: String,
/// Sets the S3 bucket name to use.
#[clap(long, env = MEILI_S3_BUCKET)]
pub s3_bucket: Option<String>,
/// Set the S3 access key. If used you must also set the secret key.
#[clap(long, env = MEILI_S3_ACCESS_KEY)]
pub s3_access_key: Option<String>,
/// Set the S3 secret key. If used you must also set the access key.
#[clap(long, env = MEILI_S3_SECRET_KEY)]
pub s3_secret_key: Option<String>,
/// Security token, can't be used with access key and secret key.
#[clap(long, env = MEILI_S3_SECURITY_TOKEN)]
pub s3_security_token: Option<String>,
/// Deactivates Meilisearch's built-in telemetry when provided.
///
/// Meilisearch automatically collects data from all instances that do not opt out using this flag.
@ -368,6 +406,13 @@ impl Opt {
http_addr,
master_key,
env,
zk_url,
s3_url,
s3_region,
s3_bucket,
s3_access_key,
s3_secret_key,
s3_security_token,
max_index_size: _,
max_task_db_size: _,
http_payload_size_limit,
@ -401,6 +446,25 @@ impl Opt {
export_to_env_if_not_present(MEILI_MASTER_KEY, master_key);
}
export_to_env_if_not_present(MEILI_ENV, env);
if let Some(zk_url) = zk_url {
export_to_env_if_not_present(MEILI_ZK_URL, zk_url);
}
if let Some(s3_url) = s3_url {
export_to_env_if_not_present(MEILI_S3_URL, s3_url);
}
export_to_env_if_not_present(MEILI_S3_REGION, s3_region);
if let Some(s3_bucket) = s3_bucket {
export_to_env_if_not_present(MEILI_S3_BUCKET, s3_bucket);
}
if let Some(s3_access_key) = s3_access_key {
export_to_env_if_not_present(MEILI_S3_ACCESS_KEY, s3_access_key);
}
if let Some(s3_secret_key) = s3_secret_key {
export_to_env_if_not_present(MEILI_S3_SECRET_KEY, s3_secret_key);
}
if let Some(s3_security_token) = s3_security_token {
export_to_env_if_not_present(MEILI_S3_SECURITY_TOKEN, s3_security_token);
}
#[cfg(all(not(debug_assertions), feature = "analytics"))]
{
export_to_env_if_not_present(MEILI_NO_ANALYTICS, no_analytics.to_string());
@ -547,7 +611,7 @@ impl TryFrom<&IndexerOpts> for IndexerConfig {
Ok(Self {
log_every_n: Some(DEFAULT_LOG_EVERY_N),
max_memory: other.max_indexing_memory.map(|b| b.get_bytes() as usize),
thread_pool: Some(thread_pool),
thread_pool: Some(Arc::new(thread_pool)),
max_positions_per_attributes: None,
skip_index_budget: other.skip_index_budget,
..Default::default()
@ -715,6 +779,10 @@ fn default_env() -> String {
DEFAULT_ENV.to_string()
}
fn default_s3_region() -> String {
DEFAULT_S3_REGION.to_string()
}
fn default_max_index_size() -> Byte {
Byte::from_bytes(INDEX_SIZE)
}

View File

@ -41,14 +41,10 @@ pub async fn create_api_key(
_req: HttpRequest,
) -> Result<HttpResponse, ResponseError> {
let v = body.into_inner();
let res = tokio::task::spawn_blocking(move || -> Result<_, AuthControllerError> {
let key = auth_controller.create_key(v)?;
Ok(KeyView::from_key(key, &auth_controller))
})
.await
.map_err(|e| ResponseError::from_msg(e.to_string(), Code::Internal))??;
let key = auth_controller.create_key(v)?;
let key = KeyView::from_key(key, &auth_controller);
Ok(HttpResponse::Created().json(res))
Ok(HttpResponse::Created().json(key))
}
#[derive(Deserr, Debug, Clone, Copy)]
@ -110,17 +106,11 @@ pub async fn patch_api_key(
) -> Result<HttpResponse, ResponseError> {
let key = path.into_inner().key;
let patch_api_key = body.into_inner();
let res = tokio::task::spawn_blocking(move || -> Result<_, AuthControllerError> {
let uid =
Uuid::parse_str(&key).or_else(|_| auth_controller.get_uid_from_encoded_key(&key))?;
let key = auth_controller.update_key(uid, patch_api_key)?;
let uid = Uuid::parse_str(&key).or_else(|_| auth_controller.get_uid_from_encoded_key(&key))?;
let key = auth_controller.update_key(uid, patch_api_key)?;
let key = KeyView::from_key(key, &auth_controller);
Ok(KeyView::from_key(key, &auth_controller))
})
.await
.map_err(|e| ResponseError::from_msg(e.to_string(), Code::Internal))??;
Ok(HttpResponse::Ok().json(res))
Ok(HttpResponse::Ok().json(key))
}
pub async fn delete_api_key(
@ -128,13 +118,8 @@ pub async fn delete_api_key(
path: web::Path<AuthParam>,
) -> Result<HttpResponse, ResponseError> {
let key = path.into_inner().key;
tokio::task::spawn_blocking(move || {
let uid =
Uuid::parse_str(&key).or_else(|_| auth_controller.get_uid_from_encoded_key(&key))?;
auth_controller.delete_key(uid)
})
.await
.map_err(|e| ResponseError::from_msg(e.to_string(), Code::Internal))??;
let uid = Uuid::parse_str(&key).or_else(|_| auth_controller.get_uid_from_encoded_key(&key))?;
auth_controller.delete_key(uid)?;
Ok(HttpResponse::NoContent().finish())
}

View File

@ -29,8 +29,7 @@ pub async fn create_dump(
keys: auth_controller.list_keys()?,
instance_uid: analytics.instance_uid().cloned(),
};
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))

View File

@ -78,6 +78,6 @@ async fn patch_features(
}),
Some(&req),
);
index_scheduler.put_runtime_features(new_features)?;
index_scheduler.inner().put_runtime_features(new_features)?;
Ok(HttpResponse::Ok().json(new_features))
}

View File

@ -1,4 +1,4 @@
use std::io::ErrorKind;
use std::io::{BufReader, ErrorKind, Seek, SeekFrom};
use actix_web::http::header::CONTENT_TYPE;
use actix_web::web::Data;
@ -129,8 +129,7 @@ pub async fn delete_document(
index_uid: index_uid.to_string(),
documents_ids: vec![document_id],
};
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))
}
@ -396,11 +395,12 @@ async fn document_addition(
return Err(MeilisearchHttpError::MissingPayload(format));
}
if let Err(e) = buffer.seek(std::io::SeekFrom::Start(0)).await {
if let Err(e) = buffer.seek(SeekFrom::Start(0)).await {
return Err(MeilisearchHttpError::Payload(ReceivePayload(Box::new(e))));
}
let read_file = buffer.into_inner().into_std().await;
let s3 = index_scheduler.s3.clone();
let documents_count = tokio::task::spawn_blocking(move || {
let documents_count = match format {
PayloadType::Json => read_json(&read_file, update_file.as_file_mut())?,
@ -409,8 +409,19 @@ async fn document_addition(
}
PayloadType::Ndjson => read_ndjson(&read_file, update_file.as_file_mut())?,
};
if let Some(s3) = s3 {
update_file.seek(SeekFrom::Start(0)).unwrap();
let mut reader = BufReader::new(&*update_file);
match s3.put_object_stream(&mut reader, format!("/update-files/{}", uuid)) {
Ok(_) | Err(s3::error::S3Error::Http(_, _)) => (),
Err(e) => panic!("Error {}", e),
}
}
// we NEED to persist the file here because we moved the `udpate_file` in another task.
update_file.persist()?;
Ok(documents_count)
})
.await;
@ -445,7 +456,7 @@ async fn document_addition(
};
let scheduler = index_scheduler.clone();
let task = match tokio::task::spawn_blocking(move || scheduler.register(task)).await? {
let task = match scheduler.register(task) {
Ok(task) => task,
Err(e) => {
index_scheduler.delete_update_file(uuid)?;
@ -476,8 +487,7 @@ pub async fn delete_documents_batch(
let task =
KindWithContent::DocumentDeletion { index_uid: index_uid.to_string(), documents_ids: ids };
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))
@ -512,8 +522,7 @@ pub async fn delete_documents_by_filter(
.map_err(|err| ResponseError::from_msg(err.message, Code::InvalidDocumentFilter))?;
let task = KindWithContent::DocumentDeletionByFilter { index_uid, filter_expr: filter };
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))
@ -529,8 +538,7 @@ pub async fn clear_all_documents(
analytics.delete_documents(DocumentDeletionKind::ClearAll, &req);
let task = KindWithContent::DocumentClear { index_uid: index_uid.to_string() };
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))

View File

@ -135,8 +135,7 @@ pub async fn create_index(
);
let task = KindWithContent::IndexCreation { index_uid: uid.to_string(), primary_key };
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
Ok(HttpResponse::Accepted().json(task))
} else {
@ -203,8 +202,7 @@ pub async fn update_index(
primary_key: body.primary_key,
};
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))
@ -216,8 +214,7 @@ pub async fn delete_index(
) -> Result<HttpResponse, ResponseError> {
let index_uid = IndexUid::try_from(index_uid.into_inner())?;
let task = KindWithContent::IndexDeletion { index_uid: index_uid.into_inner() };
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
Ok(HttpResponse::Accepted().json(task))
}

View File

@ -55,10 +55,7 @@ macro_rules! make_setting_route {
is_deletion: true,
allow_index_creation,
};
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task))
.await??
.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))
@ -97,10 +94,7 @@ macro_rules! make_setting_route {
is_deletion: false,
allow_index_creation,
};
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task))
.await??
.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))
@ -586,8 +580,7 @@ pub async fn update_all(
is_deletion: false,
allow_index_creation,
};
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))
@ -622,8 +615,7 @@ pub async fn delete_all(
is_deletion: true,
allow_index_creation,
};
let task: SummarizedTaskView =
tokio::task::spawn_blocking(move || index_scheduler.register(task)).await??.into();
let task: SummarizedTaskView = index_scheduler.register(task)?.into();
debug!("returns: {:?}", task);
Ok(HttpResponse::Accepted().json(task))

View File

@ -18,7 +18,6 @@ use serde_json::json;
use time::format_description::well_known::Rfc3339;
use time::macros::format_description;
use time::{Date, Duration, OffsetDateTime, Time};
use tokio::task;
use super::SummarizedTaskView;
use crate::analytics::Analytics;
@ -325,15 +324,12 @@ async fn cancel_tasks(
let query = params.into_query();
let (tasks, _) = index_scheduler.get_task_ids_from_authorized_indexes(
&index_scheduler.read_txn()?,
&query,
index_scheduler.filters(),
)?;
let (tasks, _) =
index_scheduler.get_task_ids_from_authorized_indexes(&query, index_scheduler.filters())?;
let task_cancelation =
KindWithContent::TaskCancelation { query: format!("?{}", req.query_string()), tasks };
let task = task::spawn_blocking(move || index_scheduler.register(task_cancelation)).await??;
let task = index_scheduler.register(task_cancelation)?;
let task: SummarizedTaskView = task.into();
Ok(HttpResponse::Ok().json(task))
@ -370,15 +366,12 @@ async fn delete_tasks(
);
let query = params.into_query();
let (tasks, _) = index_scheduler.get_task_ids_from_authorized_indexes(
&index_scheduler.read_txn()?,
&query,
index_scheduler.filters(),
)?;
let (tasks, _) =
index_scheduler.get_task_ids_from_authorized_indexes(&query, index_scheduler.filters())?;
let task_deletion =
KindWithContent::TaskDeletion { query: format!("?{}", req.query_string()), tasks };
let task = task::spawn_blocking(move || index_scheduler.register(task_deletion)).await??;
let task = index_scheduler.register(task_deletion)?;
let task: SummarizedTaskView = task.into();
Ok(HttpResponse::Ok().json(task))

View File

@ -39,7 +39,7 @@ impl Server {
let options = default_settings(dir.path());
let (index_scheduler, auth) = setup_meilisearch(&options).unwrap();
let (index_scheduler, auth) = setup_meilisearch(&options, None).await.unwrap();
let service = Service { index_scheduler, auth, options, api_key: None };
Server { service, _dir: Some(dir) }
@ -54,7 +54,7 @@ impl Server {
options.master_key = Some("MASTER_KEY".to_string());
let (index_scheduler, auth) = setup_meilisearch(&options).unwrap();
let (index_scheduler, auth) = setup_meilisearch(&options, None).await.unwrap();
let service = Service { index_scheduler, auth, options, api_key: None };
Server { service, _dir: Some(dir) }
@ -67,7 +67,7 @@ impl Server {
}
pub async fn new_with_options(options: Opt) -> Result<Self, anyhow::Error> {
let (index_scheduler, auth) = setup_meilisearch(&options)?;
let (index_scheduler, auth) = setup_meilisearch(&options, None).await?;
let service = Service { index_scheduler, auth, options, api_key: None };
Ok(Server { service, _dir: None })

View File

@ -1,3 +1,5 @@
use std::sync::Arc;
use grenad::CompressionType;
use rayon::ThreadPool;
@ -9,7 +11,7 @@ pub struct IndexerConfig {
pub max_memory: Option<usize>,
pub chunk_compression_type: CompressionType,
pub chunk_compression_level: Option<u32>,
pub thread_pool: Option<ThreadPool>,
pub thread_pool: Option<Arc<ThreadPool>>,
pub max_positions_per_attributes: Option<u32>,
pub skip_index_budget: bool,
}