Increase embedded string threshold from 64 to 128 bytes

[Nikhil-Manglore]

This PR increases the embedded string threshold to 128 bytes (2 cache lines) in order to improve memory overhead.

Note: I'll fix the tests after the benchmarks run

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 74.70%. Comparing base (3c46f66) to head (9d808ac).

Additional details and impacted files

@@             Coverage Diff              @@
##           unstable    #3397      +/-   ##
============================================
- Coverage     74.93%   74.70%   -0.24%     
============================================
  Files           130      130              
  Lines         72769    72768       -1     
============================================
- Hits          54532    54363     -169     
- Misses        18237    18405     +168     

Files with missing lines	Coverage Δ
src/object.c	91.86% <100.00%> (+0.37%)	⬆️

... and 19 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[github-actions]

Benchmark ran on this commit: e38f9c1

RPS Benchmark Comparison: unstable vs 16d2c42

2 significant change(s)

• ✅ +31±2% GET rps pipe=10 threads=9 (data_size=96)
• ✅ +5.4±0.7% GET rps pipe=10 threads=1 (data_size=96)

14 with no significant change

Click to expand full comparison tables

data_size = 16

	% Change	Test	unstable	16d2c42	unstable stats	16d2c42 stats
➖	+1.4±0.8%	GET rps P1 T1	227K	230K	n=3, σ=1.51K, CV=0.7%, CI99%=±3.8%, PI99%=±7.6%	n=3, σ=1.09K, CV=0.5%, CI99%=±2.7%, PI99%=±5.4%
➖	-0±2%	GET rps P1 T9	1.53M	1.53M	n=3, σ=28.2K, CV=1.8%, CI99%=±10.6%, PI99%=±21.2%	n=3, σ=19.4K, CV=1.3%, CI99%=±7.3%, PI99%=±14.5%
➖	+1.6±0.4%	GET rps P10 T1	1.233M	1.253M	n=3, σ=3.96K, CV=0.3%, CI99%=±1.8%, PI99%=±3.7%	n=3, σ=2.49K, CV=0.2%, CI99%=±1.1%, PI99%=±2.3%
➖	+1±1%	GET rps P10 T9	2.86M	2.896M	n=3, σ=29.7K, CV=1.0%, CI99%=±5.9%, PI99%=±11.9%	n=3, σ=7.14K, CV=0.2%, CI99%=±1.4%, PI99%=±2.8%
➖	+0.3±0.9%	SET rps P1 T1	221K	221.5K	n=3, σ=1.98K, CV=0.9%, CI99%=±5.1%, PI99%=±10.3%	n=3, σ=358, CV=0.2%, CI99%=±0.9%, PI99%=±1.9%
➖	+0.6±0.9%	SET rps P1 T9	1.483M	1.491M	n=3, σ=9.15K, CV=0.6%, CI99%=±3.5%, PI99%=±7.1%	n=3, σ=9.86K, CV=0.7%, CI99%=±3.8%, PI99%=±7.6%
➖	+1.8±0.9%	SET rps P10 T1	1.038M	1.057M	n=3, σ=8.76K, CV=0.8%, CI99%=±4.8%, PI99%=±9.7%	n=3, σ=3.02K, CV=0.3%, CI99%=±1.6%, PI99%=±3.3%
➖	+3.7±0.8%	SET rps P10 T9	1.93M	2.006M	n=3, σ=14.8K, CV=0.8%, CI99%=±4.4%, PI99%=±8.8%	n=3, σ=5.21K, CV=0.3%, CI99%=±1.5%, PI99%=±3.0%

data_size = 96

	% Change	Test	unstable	16d2c42	unstable stats	16d2c42 stats
➖	+3±1%	GET rps P1 T1	221K	227K	n=3, σ=2.01K, CV=0.9%, CI99%=±5.2%, PI99%=±10.5%	n=3, σ=2.16K, CV=1.0%, CI99%=±5.5%, PI99%=±10.9%
➖	+3±1%	GET rps P1 T9	1.470M	1.51M	n=3, σ=8.00K, CV=0.5%, CI99%=±3.1%, PI99%=±6.2%	n=3, σ=15.1K, CV=1.0%, CI99%=±5.7%, PI99%=±11.4%
✅	+5.4±0.7%	GET rps P10 T1	1.165M	1.228M	n=3, σ=3.47K, CV=0.3%, CI99%=±1.7%, PI99%=±3.4%	n=3, σ=6.87K, CV=0.6%, CI99%=±3.2%, PI99%=±6.4%
✅	+31±2%	GET rps P10 T9	2.26M	2.95M	n=3, σ=21.6K, CV=1.0%, CI99%=±5.5%, PI99%=±11.0%	n=3, σ=26.7K, CV=0.9%, CI99%=±5.2%, PI99%=±10.4%
➖	+4±1%	SET rps P1 T1	213K	220.4K	n=3, σ=1.96K, CV=0.9%, CI99%=±5.3%, PI99%=±10.5%	n=3, σ=885, CV=0.4%, CI99%=±2.3%, PI99%=±4.6%
➖	+2.1±0.7%	SET rps P1 T9	1.458M	1.49M	n=3, σ=1.30K, CV=0.1%, CI99%=±0.5%, PI99%=±1.0%	n=3, σ=10.3K, CV=0.7%, CI99%=±4.0%, PI99%=±7.9%
➖	+1.8±0.8%	SET rps P10 T1	1.025M	1.044M	n=3, σ=7.80K, CV=0.8%, CI99%=±4.4%, PI99%=±8.7%	n=3, σ=2.85K, CV=0.3%, CI99%=±1.6%, PI99%=±3.1%
➖	+8±2%	SET rps P10 T9	1.86M	2.01M	n=3, σ=34.8K, CV=1.9%, CI99%=±10.7%, PI99%=±21.5%	n=3, σ=23.6K, CV=1.2%, CI99%=±6.7%, PI99%=±13.5%

---

Configuration:

• architecture: aarch64
• benchmark_mode: duration
• clients: 1600
• cluster_mode: False
• duration: 180
• tls: False
• valkey_benchmark_threads: 90
• warmup: 30

Legend:

• Test column: Command, metric, P=pipeline depth, T=io-threads
• Significance: ✅ significant improvement, ❌ significant regression, ➖ not significant, ❔ insufficient data

Statistical Notes:

• CV: Coefficient of Variation - relative variability (σ/μ × 100%)
• CI99%: 99% Confidence Interval - range where the true population mean is likely to fall
• PI99%: 99% Prediction Interval - range where a single future observation is likely to fall