Commit Graph

14 Commits (5d66cd0a81d2533ee174e1bfb56bab86f81c1fed)

Author SHA1 Message Date
Andrew Lamb 833c10c083
feat: return write_token from HTTP writes to router2 (#4202)
* feat: return write_token from HTTP writes to router2

* fix: Update router2/src/dml_handlers/instrumentation.rs

Co-authored-by: Dom <dom@itsallbroken.com>

* refactor: Use WriteSummary::default more vigorously

* fix: fix typo and add links to follow on issues

Co-authored-by: Dom <dom@itsallbroken.com>
Co-authored-by: kodiakhq[bot] <49736102+kodiakhq[bot]@users.noreply.github.com>
2022-04-02 10:34:51 +00:00
Dom Dwyer 91730d6a1c feat: schema validation conflict/limit metrics
Emit metric counters tracking the number of schema conflicts, and number
of service limit errors observed.
2022-03-30 11:42:58 +00:00
Dom Dwyer 7c5ba34d44 refactor: enable gRPC handler
Plumbs the gRPC write handler into the existing router2 server.
2022-03-04 14:51:43 +00:00
Dom Dwyer 14d90d1011 feat: schema validation benchmarks
Useful for confirming the scalability of the schema check algorithm.
2022-03-03 23:40:13 +00:00
Dom Dwyer bb9b140f4b refactor: sequencer metrics
Records per-sequencer (kafka partition) enqueue latency / counts broken
down by operation success/error.
2022-03-03 23:40:13 +00:00
Carol (Nichols || Goulding) 3f2a58b47f
refactor: pub use data_types from data_types2
So it's clearer which parts of data_types the NG design is using, and
which types can be cleaned up eventually.
2022-03-02 13:55:31 -05:00
Dom Dwyer bd64f55658 feat: http ingest metrics
Records LP line count, field count & request body size (decompressed,
byte size) for writes, and request body byte size for deletes.
2022-03-02 13:05:55 +00:00
Marco Neumann 48722783f9
feat: offer metrics for in-mem catalog (#3876)
This can be quite helpful to test certain caching behavior w/o writing
yet-another abstraction layer.
2022-03-01 11:33:54 +00:00
Dom Dwyer d7eda88581 refactor: early schema validation
Changes the configuration of the router request pipeline to move schema
validation before partitioning.

This reduces the concurrency of callsm into the schema validator when a
single write is split into one or more partitions, reducing contention
and cash thrashing. It also ensures we don't bother partitioning the
writes if the request will fail.
2022-02-23 18:59:14 +00:00
Dom Dwyer bb132b61ad refactor: chain DML handlers
The router is composed of several DML handlers called in sequence in
order to construct the full request handling pipeline. Prior to this
commit, each handler nested the next handler it calls internally,
producing a nested call chain that resulted metrics (added in #3764)
recording cumulative latency like this:

              ┌ ─

              │     ┌───────────────┐
                    │  NS Creation  │
              │     └───────────────┘
                            │  ┌───────────────┐
              │             │  │  Partitioner  │
                            │  └───────────────┘
              │             │          │
                            │          │
 Cumulative   │             │          │  ┌───────────────┐
   Timings                1.5s        1s  │    etc...     │
              │             │          │  └───────────────┘
                            │          │
              │             │          │
                            │  ┌───────────────┐
              │             │  │  Partitioner  │
                            │  └───────────────┘
              │     ┌───────────────┐
                    │  NS Creation  │
              │     └───────────────┘

              └ ─

This meant it was hard to determine the latency of a single handler
without knowing (and subtracting the latency of) all the child handlers
it calls.

This commit replaces the intrusive nested handler call chain with an
external Chain combinator type to compose together individual handlers,
resulting in correct per-handler timings and simpler code/tests:

          ┌───────────────┐
          │  NS Creation  │
          └───────────────┘
                  │
                 .5s       ┌───────────────┐
                  └───────▶│  Partitioner  │
                           └───────────────┘
                                   │
                                  1s    ┌───────────────┐
                                   └───▶│    etc...     │
                                        └───────────────┘
2022-02-18 14:19:53 +00:00
Dom Dwyer e055800039 refactor: enable Partitioner in request pipeline
Adds the Partitioner DML handler into the handler stack, modifying the
input types of down-stream handlers to accept the partitioned data.
2022-02-15 11:34:33 +00:00
Dom Dwyer 026a557c0b refactor: rename TableNamespaceSharder
Rename to JumpHash and expose the hashing internals for reuse (outside
of only table & namespace sharding).
2022-02-04 15:56:09 +00:00
Luke Bond 011b297f28
feat: more benchmarks of router2 (#3575) 2022-01-28 17:44:10 +00:00
Luke Bond 4a96e52290
feat: router2 sharder benchmarking (#3558)
* feat: benchmarking the router2 sharder

* chore: added throughput to sharder benchmarks; vary num buckets
2022-01-27 18:09:16 +00:00