| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
commit_hash:109d7b77afce712c68e57058c5f38e52eade292e
|
| |
|
|
| |
commit_hash:2a15f45c98aa40190e39522a2155d9ddcf83263b
|
| |
|
|
| |
commit_hash:bbd5ab6a8913a7d713da2082152a643532831810
|
| |
|
|
|
|
|
|
|
|
|
|
| |
tablet writes
* Changelog entry
Type: feature
Component: dynamic-tables
All prepare signatures are now equal to 1. This enables re-assembling commits across cell commit
sessions on write retry, e.g. when a tablet moves.
commit_hash:40ce0e6541a09888417f5b84b25012b796c4bfdd
|
| |
|
|
|
|
|
|
|
|
|
| |
Also affects all schemaful pipes - backpressure and stream fragmentation are implemented.
* Changelog entry
Type: feature
Component: dynamic-tables
Implemented parallel group by on proxy, significantly improving latency of high-cardinality grouping queries.
commit_hash:29af1e557205b21e22268a1fa333172a9d1fc681
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: offshore data gateway
This PR adds the `OffshoreDataGateway` component which is used when reading data from S3. This component is also integrated into the replication reader to actually perform the reads.
One of the most important topics to note in this work is the introduction of the new proto message `TChunkReplicaSpec` and its usage in the `OffshoreDataGateway`'s RPC requests. This structure allows us to pass the medium index which is required to find the specific offshore medium where the chunk is located. Maybe there are other solutions, I'll be happy to discuss that.
Another topic is the approach to handling different media in the replication reader. Right now the replication reader uses only addresses to uniquely identify different peers to read from. This works because we read from data nodes only, and we do not care about the medium there. Now the reads may also happen from `OffshoreDataGateway`-s, and the medium starts to matter.
Imagine a scenario when a chunk has two offshore replicas on two different mediums, one pointing to Google S3 storage, another to AWS S3. We must have a way to differentiate those two replicas, even though the address is the same - the sentinel `OffshoreNodeAddress`. This is why I introduce a structure called `TPeerId` which includes both the address and the medium index, and now replication reader works with it to differentiate different replicas.
Last topic is the testing. It's impossible to implement an "honest" integration test at the moment as writes to offshore replicas are not implemented, and also masters know nothing of them. I have implemented a C\+\+ unit test to check the behaviour of replication reader only - `test_s3_data.cpp`.
---
Pull Request resolved: <https://github.com/ytsaurus/ytsaurus/pull/1688>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
Co-authored-by: cherepashka <[email protected]>
commit_hash:7941b82f5735c5788fefec1ccf5175ddd86528a5
|
| |
|
|
| |
commit_hash:8d9c3b0bbddd5c1c725cc3c1eefddcc6498f936a
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: proxy
Introduce "FetchCookieNodeDescriptors" flag to serialize node directory for partition inside cookie. It should prevent unresolved node id errors during ReadTablePartition call.
commit_hash:fbdda26266e4efec8f1202c5ccdcd98b905db928
|
| |
|
|
|
| |
<https://nda.ya.ru/t/lADBX2ru7ZPHnh>
commit_hash:3c379194a4882d7ed93f922d5bf5ea25e6ee0ddd
|
| |
|
|
|
|
|
|
| |
* Changelog entry
Type: fix
Component: dynamic-tables
Write, Delete, Lookup Rows allowed missing keys if allowMissingKeyColumns flag is set
commit_hash:4f90a0f906fab5ad43914a34c4737f1f78986a3e
|
| |
|
|
|
|
| |
node config
commit_hash:9fb73ada8d5eb337a1f9fd0962b589e7ff21fe76
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: api
Support compressed data size in PartitionTables
commit_hash:ff707e5628ee3096ac17f0a3f04cbc1a92916996
|
| |
|
|
|
|
|
|
| |
columnar statistics
'estimation' refers to the process; 'estimate' is the result value.
Rename the proto field and output key accordingly.
commit_hash:9654e9efe2b2d73f49b4ce22b0aa60dbff5eeece
|
| |
|
|
|
|
| |
cell_directory.cpp into helpers.cpp
commit_hash:ae257311dc7104cf31e35dd26c1c41ddfda4353d
|
| |
|
|
| |
commit_hash:8ed3cffde90496b2de8440d5b59674301761de28
|
| |
|
|
|
|
|
|
|
|
|
| |
* Changelog entry
Type: fix
Component: misc-api
Switch code field from required to optional.
<Message for release notes>
commit_hash:e86cf4ef3843f8f0bbd36267dd1a0298f8b203b7
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: proxy
Worker pool in rpc proxies is made fair-share.
commit_hash:9ce485656813b6821badbf1bc855ce79809b7206
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: master & node
Added location index into chunk confirmation
commit_hash:1f0ba59fdbe8f25bf80086b312d8df17e6a3f464
|
| |
|
|
|
|
|
|
|
|
|
| |
* Changelog entry
Type: fix
Component: proxy
Rename shadowed data member NYT::TStartDistributedWriteTable(File)Options::Timeout -\> SessionTimeout
<Message for release notes>
commit_hash:1b5b967b5136a7105627854044de6eee634dde03
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: dynamic-tables
Unfolded columns in unfolding indices are allowed to have different names in table and in index.
commit_hash:e4761712a3b4cd69f215f69ea496118b9eff6990
|
| |
|
|
| |
commit_hash:e8801b81ac480cb7b5c883c912d4ec2d6f3c74d7
|
| |
|
|
|
|
|
|
|
|
|
| |
cluster node registration and heartbeat
* Changelog entry
Type: feature
Component: master & node
Made masters to return secondary masters configs as a response to cluster node registration and heartbeat
commit_hash:5136cda005902249cf2feafcf205e289f0ee1235
|
| |
|
|
| |
commit_hash:8cfcaa94f1147884e6d5c2a91e2728889fc906cd
|
| |
|
|
| |
commit_hash:14d1302c7ee0b288f4f5f1625abcb0e2da0b59d1
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: proxy
Worker pool in rpc proxies is made fair-share.
commit_hash:7e3b72360ee72775a46b7812ea54cfa1ca3c703e
|
| |
|
|
| |
commit_hash:217465a241bdd9018ca30954e645ffa9917ffd6c
|
| |
|
|
|
|
| |
* Do not wait for the prepared transaction barrier on `yt <command> --params '{suppress_strongly_ordered_transaction_barrier=%true}'`.
* Make HTTP & Cypress Proxies registration commands ignore the barrier.
commit_hash:52ffad5a6418fd9d2cb1af763b33fc1b19f2a648
|
| |
|
|
| |
commit_hash:aa3a56ee71eec7893d8e7986d35eed908a083f5f
|
| |
|
|
| |
commit_hash:0646bf50b5be7e93f78eb08ed151859116a8a7c3
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: scheduler
Introduce run-job-shell-command: API to run command in job shell and receive its output.
commit_hash:1c0204a7548195d945b0915e5ff734ea48438532
|
| |
|
|
| |
commit_hash:71a0cdcaf571a4ed247dfe29e9a5f0850a2a9c5a
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
method
<Message NOT for release notes>
* Changelog entry
Type: feature
Component: proxy
This change adds read size estimation support to the GetColumnarStatistics RPC method. Read size estimation provides a good approximation of compressed data size for column groups when chunks have the same column groups as the table schema.
<Message for release notes>
commit_hash:43c2dfda01b2b274f24beb43f1bef7576c87f276
|
| |
|
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: dynamic-tables
Introduce transfer_bundle_resources command. It allows to transfer resources
between two tablet cell bundle atomically.
commit_hash:14ac71963fe8594c4c4d7f200812ca06fcb2a0b6
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: master cpp-sdk
1. Optionally allow renaming and removing fields of structs and variant structs.
2. Provide separate options for static and dynamic tables.
3. Add unit- and integration tests.
4. Support in chunk readers will be implemented separately.
commit_hash:3d5d4412381b5b35f90ba81488836f9f40a04ff7
|
| |
|
|
| |
commit_hash:c1cf6352dd53436db248717a38c82f686a3f600c
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: master
Introduce constraints for columns in table node schema on master side
API:
Create and Alter Table commands will have 2 more parameters: `constrained_schema` and `constraints`. The first parameter represents usual schema with columns marked by constraints, for example: `[{"constraint": "BETWEEN 0 AND 60", "name": "minute", "stable_name": "_minute", "type": "int64", "sort_order": "ascending"}, {"constraint": "BETWEEN 0 AND 23", "name": "hour", "type": "int64"}]`. The second parameter represents mapping between name of column into constraint, for example: `{"minute": "BETWEEN 0 AND 60", "hour": "BETWEEN 0 AND 23"}`. Addition of `constrained_schema` allows us to save the old behavior with altering tables and avoid inadvertent removals of constraints which would inevitably arise from mixing up new (constained) and old (unconstrained) schemas. Addition of `constraints` parameter will allow users to add constraints even if they didn't want to change schema of table.
Tables will have 2 more attributes: `@constrained_schema` and `@constraints`. They are just similar to new parameters in create and alter-tables commands.
Deduplication of schemas will work as before: schemas with no constraints will be deduplicated by `@schema_id`. Constraints from schemas will not be deduplicated yet (by `@constraints_id`), this feature will be supported in future PRs.
commit_hash:eda85e01afc0e89398196327c5b5de7f334e96c0
|
| |
|
|
| |
commit_hash:cfd8cb99e2ef5ffe758f397dbb2180cfbb795222
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: scheduler
Introduce check\_operation\_permission: API to verify user's permission to operation.
commit_hash:5dd48763fd15e5a912351f959f8d120e6b3f140e
|
| |
|
|
| |
commit_hash:e0f1278fe0e37eca9ebbd11456f617c3665b4b32
|
| |
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: scheduler
Introduce list-job-traces: API to receive info about job's GPU traces
commit_hash:0587ef97d31238f84bb3b1f69dc17e1f202610a0
|
| |
|
|
| |
commit_hash:c2fc4993b3226aacc10fee87fef584f8a5360a88
|
| |
|
|
| |
commit_hash:d5a8edf11a5436650f288d4cd884480e778a6387
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: fix
Component: dynamic-tables
Compaction timestamp digest fix for min\_data\_versions=1.
commit_hash:9bd762943f60d9e82be30cb698b574fd45e79049
|
| |
|
|
|
|
|
|
|
| |
* Changelog entry
Type: feature
Component: proxy
Introduce distributed write protocol for files
commit_hash:6e90fa9dc228b4b1b9fecc19473152f851c39ce1
|
| |
|
|
|
|
|
|
|
|
|
| |
<Message NOT for release notes>
* Changelog entry
Type: fix
Component: scheduler
Rework get\_job\_trace api: enable streaming and return perfetto-friendly format
commit_hash:082e255afb9022eb137dac3d0c3be8ac49d3b5e2
|
| |
|
|
| |
commit_hash:7993a8aa42588e32d89c7620cfdbce19bb5e4cde
|
| |
|
|
| |
commit_hash:0a2dda0b5dc87a8fccae9081a099d1fc2f1d06e1
|
| |
|
|
| |
commit_hash:0c4facfeb34aaab4f733073b297329fa923702a4
|
| |
|
|
| |
commit_hash:db18c78c543df7c113c96252143bfd877f57e215
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Makes it possible to run map operation job groups consisting of several containers, distinguishing one from another using `YT_JOB_COOKIE_GROUP_INDEX` env-variable.
Only the main job (`YT_JOB_COOKIE_GROUP_INDEX == 0`) receives the input and output streams.
Group elements can be queried using the `main_job_id` filter.
`YT_JOB_COOKIE_MAIN_JOB_ID` lets the jobs find the main job.
https://docs.google.com/document/d/13eXf7i2ylDybhasRoocZlgdpOk0ttCXjOKIC5xGHGdc/edit?tab=t.0#heading=h.8kt1zjdp11cm
---
Pull Request resolved: https://github.com/ytsaurus/ytsaurus/pull/1149
Co-authored-by: pogorelov <[email protected]>
Co-authored-by: pogorelov <[email protected]>
Co-authored-by: pogorelov <[email protected]>
Co-authored-by: pogorelov <[email protected]>
Co-authored-by: pogorelov <[email protected]>
Co-authored-by: pogorelov <[email protected]>
Co-authored-by: pogorelov <[email protected]>
Co-authored-by: pogorelov <[email protected]>
commit_hash:edf43e210f9db0974047bba3084777cc07d51e89
|