* feat: rework cache refresh logic
Instead of issuing a single refresh when a GET request for a cached key
comes in, start a background job (using some efficient logic to not
overload tokio) per key that refreshes the key using some exponential
backoff. The timer is reset a new GET request comes in. This has the
following advantages:
- our backoff logic decorrelates the requests
- the longer a key was not used, the less often it will be updated
All test (esp. integration tests) as adjusted accordingly, mostly to
account for the fact that no extra GET is required to start the refresh
timer.
Closes#5720.
* docs: improve
Co-authored-by: Andrew Lamb <alamb@influxdata.com>
* refactor: simplify rng overwrite
Co-authored-by: Andrew Lamb <alamb@influxdata.com>
Use a proper typed stream instead of peeking the first element. This is
more in line with our remaining stack and shall also improve error
handling.
Co-authored-by: kodiakhq[bot] <49736102+kodiakhq[bot]@users.noreply.github.com>
* feat: Reimplement keywords as BTreeSet for more efficient lookup
Added a `Token` type to perform case-insensitive comparisons to avoid
allocations when performing lookups in the set.
Also introduced new `keyword` combinator to replace use of
`tag_no_case`. This is necessary to prevent eagerly recognising tags
without the appropriate separator. For example, `tag_no_case("OR")`
will recognise `ORDER` in a conditional expression.
The new combinator a `keyword` follows the pattern:
```
keyword ::= (a..z | A..Z)* keyword_separator
keyword_separator ::= ' ' | '(' | ')' | ';' | ',' | '=' |
'\n' | '\t'
```
* fix: Allow quoted time; using more efficient case-insensitive comparison
* fix: Digits to right of decimal point are optional
* fix: More idiomatic use of zip
* fix: Use `keyword` combinator and adjust whitespace handling
* fix: <> is a valid alias for the != operator
* fix: Special handling of DISTINCT identifier for function call name
* chore: Add tests
* chore: Feedback to remove unnecessary comments
* chore: Switch from BTreeSet to HashSet – thanks @domodwyer!
* fix: Avoid some allocations by collecting instead of inserting into a vec
* refactor: Encode that adding columns is for one table at a time
* test: Add another test of column limits
* test: Add below/above limit tests for create_or_get_many
* fix: Explicitly DO NOT check column limits when inserting many columns
* feat: Cache the max_columns_per_table on the NamespaceSchema
* feat: Add a function to validate column limits in-memory
* fix: Provide more useful information when over column limits
* fix: Swap types to remove intermediate allocation
* docs: Explain the interactions of the cache and the column limits
* test: Actually set up test that showcases column limit race condition
* fix: Allow writing to existing columns even if table is over column limit
Co-authored-by: Dom <dom@itsallbroken.com>
* feat: Partition implementation of Visitable for InfluxQL AST
* feat: Added consistent structures for each clause to simplify visitor
Continued to expand `accept` and `pre` / `post` visit implementations.
* feat: Added insta and tests using snapshots (thanks @crepererum)
The insta crate simplifies the process of validating the combination of
visitor and accept implementations are called and in the correct order.
* chore: Run cargo hakari tasks
* feat: Added remaining snapshot tests
Some tests are failing as some minor type changes must be added along
with the addition of related visitor functions.
* feat: Add types to represent each clause in numerous statements
These clauses permit distinct visit functions on the `Visitor` type.
* chore: Reformat `SELECT`
* chore: Explicitly specify access to export selected types only
This required completing all the missing documentation for the exported
types.
* chore: Update Cargo.lock
* chore: macro to implement common traits and hide 0th tuple element
Co-authored-by: CircleCI[bot] <circleci@influxdata.com>
* revert: "revert: rdkafka/rskafka swapping (#5800)"
This reverts commit b77c3540e1.
* test: Verify write buffer connection_config is parsed as expected
* test: Failing test reproducing the error seen when deploying rdkafka
* fix: Translate k8s-idpe configs to rdkafka configs
* feat: Add back rdkafka dependency
* feat: Remove RSKafkaProducer
* feat: Remove write buffer RecordAggregator
* feat: Add back rdkafka producer
Using code from 58a2a0b9c8311303c796495db4f167c99a2ea3aa then getting it
to compile with the latest
* feat: Add a metric around enqueue
* fix: Remove unused imports
* fix: Increase Kafka timeout to 20s
* docs: Clarify that Kafka topics should only be created in test/dev envs
* fix: Remove metrics that aren't needed for this experiment
Co-authored-by: Dom <dom@itsallbroken.com>
Co-authored-by: kodiakhq[bot] <49736102+kodiakhq[bot]@users.noreply.github.com>
Adds a new DeferredSortKey type that fetches a partition's sort key from
the catalog in the background, or on-demand if not yet pre-fetched.
From the caller's perspective, little has changed compared to reading it
from the catalog directly - the sort key is always returned when calling
get(), regardless of the mechanism, and retries are handled
transparently. Internally the sort key MAY have been pre-fetched in the
background between the DeferredSortKey being initialised, and the call
to get().
The background task waits a (uniformly) random duration of time before
issuing the catalog query to pre-fetch the sort key. This allows large
numbers of DeferredSortKey to (randomly) smear the lookup queries over a
large duration of time. This allows a large number of DeferredSortKey to
be initialised in a short period of time, without creating an equally
large spike in queries against the catalog in the same time period.
* feat: Upload in small chunks and in parallel
* fix: doclink
* fix: Apply suggestions from code review
Co-authored-by: Carol (Nichols || Goulding) <193874+carols10cents@users.noreply.github.com>
* fix: Update influxdb_iox_client/src/client/write.rs
* fix: fixup error handling and fmt
* fix: Make default chunk sizes the same and add docs
* fix: clippy
Co-authored-by: Carol (Nichols || Goulding) <193874+carols10cents@users.noreply.github.com>
* feat: Identifier → VarRef, which represents a tag or field reference
InfluxQL calls these variables, so the Rust parser will follow
precedent.
* chore: Add docs
* chore: Separate arithmetic and conditional expression parsing
This is necessary to ensure conditional expressions are only supported
in a `WHERE` clause and arithmetic expressions in `SELECT` field
projection list
* feat: Add a `verify` API to transform combinator functions to errors
This is useful to preform additional validation on a valid result.
* feat: Add customisation for operand parsing to arithmetic expressions
This permits narrowing of valid function calls in `WHERE` clauses
and to customise valid operands in a `SELECT` projection, such as a
wildcard (`*`)
* chore: move logic to appropriate modules.
* feat: Add wildcard discriminator to Expr enum required for `SELECT`
* chore: Remove dead_code attribute
* feat: Add `DISTINCT` operator
* chore: Add `distinct` macro; simplify existing macros
* chore: add assert_matches; parse Field and Field list; refactor literal
* feat: Add a new type to parse a separated_list1 into a specialised type
* chore: Refactor types using `separated_list1`
* chore: Ensure items in a list can be preceded by a space
* chore: Refactor IN clause to use OneOrMore
* feat: Parse `GROUP BY` clause
* chore: appease clippy
* chore: refactor number parsing to share code; add signed Number type
* feat: completed `SELECT` statement parser.
* chore: Consistent error messages when expecting tokens
* chore: Add recommended derive implementations
* feat: Add `SELECT` statement to `statement` combinator
* chore: Refactor OneOrMore::separated_list1 to a single error message
* chore: More tests and cleanup 🧹
* chore: More tests
* chore: Appease broken doc links checker
* chore: remove remaining `dead_code` attributes 🥳
* chore: Address PR feedback
* https://github.com/influxdata/influxdb_iox/pull/5692#discussion_r982559420
* https://github.com/influxdata/influxdb_iox/pull/5692#discussion_r982527476
* chore: Address PR feedback
* https://github.com/influxdata/influxdb_iox/pull/5692#discussion_r982626857
* https://github.com/influxdata/influxdb_iox/pull/5692#discussion_r982569760
* chore: Address PR feedback for Display implementations
Also removed duplicate test case
Marco Neumann
dependabot[bot]
Andrew Lamb
Dom Dwyer
Stuart Carnie
Carol (Nichols || Goulding)
kodiakhq[bot]
Andrew Lamb
Jake Goulding