-
Notifications
You must be signed in to change notification settings - Fork 0
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
design: the bucket index is authoritative, and that is the root of the multi-writer defects
enhancementNew feature or requestNew feature or requestquestionFurther information is requestedFurther information is requestedStatus: Open.#415 In oteldb/storage;flaky: TestCacheOversizeDoesNotDropWorkingSet asserts an exact read count against an async admission policy
flaky-testIntermittently failing testIntermittently failing testStatus: Open.#410 In oteldb/storage;partsync: the bucket index mirror writes unconditionally, safe only because partsync implies a private backend
enhancementNew feature or requestNew feature or requestStatus: Open.#399 In oteldb/storage;cluster/router: an embedded etcd per parallel test makes read_test flaky on Windows
flaky-testIntermittently failing testIntermittently failing testStatus: Open.#391 In oteldb/storage;cluster/partsync: nothing repairs an owner that lost a part — the replica keeps the last copy and says nothing
data-lossRecords or metadata can be silently and permanently lostRecords or metadata can be silently and permanently lostenhancementNew feature or requestNew feature or requestStatus: Open.#387 In oteldb/storage;cluster: peer transport and etcd have no TLS or authentication
enhancementNew feature or requestNew feature or requestsecurityAuthentication, authorization, transport securityAuthentication, authorization, transport securityStatus: Open.#374 In oteldb/storage;perf(recordengine): a fetch resolves each part's streams three times, then throws the answer away
performanceThroughput, latency, memory or I/O efficiencyThroughput, latency, memory or I/O efficiencyStatus: Open.#365 In oteldb/storage;perf(block): one self-encoded granule costs 4-13x on full-column decode — sharedDictMinRepeat trades full scans against windowed reads and knows about neither
performanceThroughput, latency, memory or I/O efficiencyThroughput, latency, memory or I/O efficiencyStatus: Open.#358 In oteldb/storage;perf(logs): choose the body codec per part — dict is 3x better on repetitive parts, raw 4-17% better on near-unique ones
performanceThroughput, latency, memory or I/O efficiencyThroughput, latency, memory or I/O efficiencyStatus: Open.#329 In oteldb/storage;perf(recordengine): merge rebuilds what it was already handed (umbrella)
performanceThroughput, latency, memory or I/O efficiencyThroughput, latency, memory or I/O efficiencyStatus: Open.#328 In oteldb/storage;benchmark CI groups head and base samples, so runner drift reads as a significant regression
flaky-testIntermittently failing testIntermittently failing testStatus: Open.#316 In oteldb/storage;Legacy parts strand permanently: seal test compares an uncompressed estimate against the compressed merge cap
bugSomething isn't workingSomething isn't workingStatus: Open.#297 In oteldb/storage;