Issues

• Setup-only cluster reconnects shorten partial-run duration and inflate rates

  redis/memtier_benchmark#819 · Filter by author fcostaoliveira opened on Oct 23, 2026

• Deadline-clamped reconnect uses a fresh clock against libevent cached time and can start a client late

  redis/memtier_benchmark#811 · Filter by author fcostaoliveira opened on Oct 23, 2026

• Flaky tests seen in CI (prometheus latency cross-check, CLIENT KILL timing, injector compile timeout, ASK reconnect timeout)

  redis/memtier_benchmark#740 · Filter by author fcostaoliveira opened on Oct 16, 2026
  1

• --reconnect-on-error with --requests and unlimited --max-reconnect-attempts: a connection whose setup keeps being rejected reconnects forever

  redis/memtier_benchmark#693 · Filter by author fcostaoliveira opened on Oct 16, 2026

• Cluster --read-preference secondary: an idle connection to a primary that dies is never noticed, so the slot map is never refreshed

  redis/memtier_benchmark#689 · Filter by author fcostaoliveira opened on Oct 16, 2026

• Cluster: requests for a slot with no owner in a valid CLUSTER SLOTS reply are never sent; the thread spins and the run ignores --test-time/--requests

  redis/memtier_benchmark#685 · Filter by author fcostaoliveira opened on Oct 15, 2026

• Cluster --read-preference secondary: read-only runs never refresh the slot map, so reads keep going to a replica promoted by CLUSTER FAILOVER (intended?)

  redis/memtier_benchmark#678 · Filter by author fcostaoliveira opened on Oct 15, 2026

• Cluster --reconnect-on-error: when the seed node (-s/-p) dies, the whole client stops even after its replica is promoted

  redis/memtier_benchmark#677 · Filter by author fcostaoliveira opened on Oct 15, 2026

• Cluster mode: keyless --command requests are sent on replica connections even with --read-preference primary (or no --read-preference)

  redis/memtier_benchmark#676 · Filter by author fcostaoliveira opened on Oct 15, 2026

• Cluster --read-preference: a write sent to a replica gets -MOVED to the slot's own primary, and every such reply triggers a CLUSTER SLOTS refresh

  redis/memtier_benchmark#664 · Filter by author fcostaoliveira opened on Oct 15, 2026

• Cluster mode without --retry-on-error: a -MOVED/-ASK reply is counted as a completed op (Count > commands the server ran) - intended?

  redis/memtier_benchmark#650 · Filter by author fcostaoliveira opened on Oct 14, 2026

• Cluster --transaction: 'pin connection received MOVED mid-rotation' warning is printed once per rotation during a slot migration (hundreds of lines)

  redis/memtier_benchmark#649 · Filter by author fcostaoliveira opened on Oct 14, 2026

• --transaction with --requests: run can end after a MULTI without its EXEC when --requests is not a multiple of the block length

  redis/memtier_benchmark#647 · Filter by author fcostaoliveira opened on Oct 14, 2026

• tests: harness crashes under RLTest --env existing-env and -s (dbDirPath / missing redis log, CONFIG GET reply sliced as a list)

  redis/memtier_benchmark#641 · Filter by author fcostaoliveira opened on Oct 14, 2026

• TLS cluster: worker stalls after reconnects (unsent output, handshake never completes)

  redis/memtier_benchmark#562 · Filter by author fcostaoliveira opened on Oct 12, 2026
  1

• libevent 2.1.12 4 KiB read cap depresses pipelined 1 KiB GET results

  redis/memtier_benchmark#535 · Filter by author gs252525 opened on Sep 16, 2026
  2

  

• Verify whether bufferevent_write is actually required for READONLY/CLIENT NO-TOUCH on discovered connections, or the original diagnosis was wrong

  redis/memtier_benchmark#532 · Filter by author fcostaoliveira opened on Aug 28, 2026
  1

• All CI test cells depend on a personal fork's availability (tests/test_requirements.txt -> fcostaoliveira/RLTest)

  redis/memtier_benchmark#530 · Filter by author fcostaoliveira opened on Aug 28, 2026

• Investigate whether peer_client_has_any_setup_in_progress() gating on is_ready_for_reads() (not is_conn_setup_done()) can livelock read routing

  redis/memtier_benchmark#529 · Filter by author fcostaoliveira opened on Aug 28, 2026

• Consider a generic --connection-setup-command flag instead of per-command booleans

  redis/memtier_benchmark#528 · Filter by author fcostaoliveira opened on Aug 27, 2026
  2

• Support CLIENT NO-TOUCH as a connection-setup command

  redis/memtier_benchmark#525 · Filter by author fcostaoliveira opened on Aug 27, 2026

• SIGABRT: m_pending_resp goes negative on multi-channel SUBSCRIBE/PSUBSCRIBE/UNSUBSCRIBE

  bug
  redis/memtier_benchmark#515 · Filter by author fcostaoliveira opened on Jul 31, 2026

• Customers commonly exhaust AWS burst credits on Elasticache. Need to measure network, cpu and iops against ec2 instance limits.

  redis/memtier_benchmark#514 · Filter by author alberttwong opened on Jul 16, 2026

• Only send HELLO when the server supports it (avoid breaking RESP2 against pre-6.0 servers)

  redis/memtier_benchmark#488 · Filter by author fcostaoliveira opened on Jun 24, 2026

• Pre-existing latent hang in zipfian arbitrary-command path (test_zipfian_arbitrary_command_mixed_operations) under UBSan

  redis/memtier_benchmark#482 · Filter by author fcostaoliveira opened on Jun 17, 2026