Skip to content

Occamy against Guacamole

The comparison benchmark's results in bench-essential.json, as bench/report.py analyses them: each series' median of every metric in each scenario and network, the improvement of one series over the other, and whether the difference is significant. How Occamy is measured describes the scenarios, the networks and the metrics, and Analysis the statistics, which were fixed before any data was collected.

The run

  • Started at 2026-11-03T20:47:56.02838Z, on Apple M5 Pro (18 CPUs, darwin), with podman version 6.0.2.
  • 6 runs with seed 20261102; each scenario that measures a key's latency timed 100 keys a run.
  • Before each result, the harness waited until the host's 1-minute load average was below 8.
  • 192 results analysed, and 0 failed.
  • The series are Guacamole and Occamy; Run a comparison gives each one's name in the results files.

How to read this report

Each value is the median of the runs, followed by its 95% interval in brackets: an exact interval from order statistics, which assumes nothing about how the values are distributed.

Improvement is the baseline's value divided by the candidate's, or the candidate's divided by the baseline's for a metric where higher is better, such as frames shown a second. Above 1×, the candidate is better: 3× is a third of the baseline's latency, or three times its frames. It's the Hodges–Lehmann estimate of that ratio, with its exact 95% interval from the Mann–Whitney distribution.

p is the exact two-sided permutation test of the rank sum, over every split of the runs. Each scenario's primary metric is the one its comparison is judged on. p (Holm) corrects the primary metrics' p-values with Holm's method, and a difference is significant where it's below 0.05. The other metrics, the secondary ones, are exploratory, and their p-values aren't corrected.

Primary metrics

Each scenario's primary metric for Guacamole and Occamy, and the improvement of Occamy over Guacamole: above 1×, Occamy is better. Holm's correction is over these 16 tests. Target says whether the improvement meets the target of 10×, which it does where its interval's lower bound is at least 10×.

Scenario Network Metric Guacamole Occamy Improvement p p (Holm) Target
Typing, SSH No delay Latency, 95th percentile (ms) 2.48 [2.20, 5.39] 1.73 [1.41, 2.63] 1.49× [1.18, 2.21] 0.0260 0.0519 no
Long output, SSH No delay Time to complete (ms) 930 [864, 978] 19.7 [16.6, 48.6] 46.8× [20.1, 55.7] 0.0022 0.0346 met
Key press, VNC No delay Latency, 95th percentile (ms) 16.4 [15.7, 17.0] 4.64 [4.28, 5.31] 3.49× [3.21, 3.73] 0.0022 0.0346 no
Key press, RDP No delay Latency, 95th percentile (ms) 20.6 [19.3, 22.4] 8.10 [6.92, 12.3] 2.55× [1.82, 2.98] 0.0022 0.0346 no
Key during motion, VNC No delay Latency, 95th percentile (ms) 155 [132, 165] 49.0 [39.6, 54.3] 3.13× [2.71, 3.87] 0.0022 0.0346 no
Key during motion, RDP No delay Latency, 95th percentile (ms) 150 [135, 191] 66.8 [57.9, 92.3] 2.27× [1.83, 2.76] 0.0022 0.0346 no
Full-screen photo, VNC No delay Data to the client (KiB) 2,235 [1,497, 2,235] 122 [122, 122] 18.3× [12.3, 18.3] 0.0022 0.0346 met
Full-screen photo, RDP No delay Data to the client (KiB) 2,175 [1,379, 2,175] 108 [108, 115] 20.1× [12.7, 20.1] 0.0022 0.0346 met
Typing, SSH 50 ms, 20 Mbit/s Latency, 95th percentile (ms) 55.4 [54.3, 55.8] 54.5 [53.4, 55.1] 1.02× [0.999, 1.03] 0.0649 0.0649 no
Long output, SSH 50 ms, 20 Mbit/s Time to complete (ms) 951 [917, 1,316] 107 [105, 109] 8.89× [8.54, 12.1] 0.0022 0.0346 no
Key press, VNC 50 ms, 20 Mbit/s Latency, 95th percentile (ms) 69.4 [67.5, 73.3] 57.5 [56.4, 59.0] 1.21× [1.18, 1.24] 0.0022 0.0346 no
Key press, RDP 50 ms, 20 Mbit/s Latency, 95th percentile (ms) 71.4 [70.2, 72.6] 60.0 [59.3, 60.7] 1.19× [1.18, 1.21] 0.0022 0.0346 no
Key during motion, VNC 50 ms, 20 Mbit/s Latency, 95th percentile (ms) 214 [182, 229] 93.2 [91.1, 113] 2.19× [1.97, 2.47] 0.0022 0.0346 no
Key during motion, RDP 50 ms, 20 Mbit/s Latency, 95th percentile (ms) 226 [201, 240] 117 [111, 126] 1.91× [1.70, 2.05] 0.0022 0.0346 no
Full-screen photo, VNC 50 ms, 20 Mbit/s Data to the client (KiB) 2,235 [1,441, 2,235] 122 [122, 122] 18.3× [11.8, 18.3] 0.0022 0.0346 met
Full-screen photo, RDP 50 ms, 20 Mbit/s Data to the client (KiB) 2,175 [1,169, 2,175] 108 [108, 111] 20.1× [10.8, 20.1] 0.0022 0.0346 met

14 of the 16 differences are significant after the correction, and 5 of the 16 improvements meet the target of 10×.

The improvement of Occamy over Guacamole on each scenario's primary metric, with its 95% interval, on a log axis, in a panel for each network; a dashed line marks the target of 10×The improvement of Occamy over Guacamole on each scenario's primary metric, with its 95% interval, on a log axis, in a panel for each network; a dashed line marks the target of 10×

Secondary metrics

The other metrics that each scenario records, for Guacamole and Occamy. They're exploratory, and their p-values aren't corrected.

Scenario Network Metric Guacamole Occamy Improvement p
Typing, SSH No delay Latency, median (ms) 1.49 [1.28, 1.67] 0.96 [0.77, 1.18] 1.54× [1.22, 1.98] 0.0022
Typing, SSH No delay Server memory, peak (MiB) 62.7 [43.6, 78.9] 32.9 [6.32, 65.6] 2.31× [0.787, 6.99] 0.1797
Typing, SSH No delay Client CPU time (ms) 63.3 [51.6, 67.9] 60.7 [53.9, 66.2] 1.02× [0.852, 1.17] 0.6991
Long output, SSH No delay Data to the client (KiB) 101 [60.4, 113] 8.95 [5.64, 9.67] 11.4× [7.60, 14.3] 0.0022
Long output, SSH No delay Server CPU time (ms) 892 [838, 926] 20.2 [15.7, 24.6] 43.8× [37.6, 55.2] 0.0022
Long output, SSH No delay Server memory, peak (MiB) 68.0 [63.2, 93.4] 32.3 [8.30, 61.8] 3.02× [1.21, 7.61] 0.0022
Long output, SSH No delay Client CPU time (ms) 47.5 [34.1, 53.1] 22.7 [20.3, 26.9] 1.97× [1.45, 2.32] 0.0022
Key press, VNC No delay Latency, median (ms) 14.6 [14.3, 14.6] 3.47 [3.37, 3.66] 4.18× [4.00, 4.24] 0.0022
Key press, VNC No delay Server memory, peak (MiB) 75.8 [73.5, 76.8] 34.6 [13.7, 62.1] 2.49× [1.24, 5.35] 0.0022
Key press, VNC No delay Client CPU time (ms) 67.2 [54.8, 70.6] 76.4 [67.1, 90.7] 0.833× [0.695, 1.01] 0.0931
Key press, RDP No delay Latency, median (ms) 17.1 [16.7, 17.8] 5.85 [5.46, 6.59] 2.90× [2.70, 3.13] 0.0022
Key press, RDP No delay Server memory, peak (MiB) 92.4 [58.3, 93.5] 44.7 [39.7, 51.4] 2.06× [1.47, 2.32] 0.0022
Key press, RDP No delay Client CPU time (ms) 72.8 [63.6, 85.3] 99.2 [82.1, 114] 0.728× [0.62, 0.854] 0.0043
Key during motion, VNC No delay Frames shown a second (frames/s) 25.2 [24.0, 26.6] 29.1 [28.4, 29.7] 1.15× [1.11, 1.22] 0.0022
Key during motion, VNC No delay Scene frames never shown (%) 16.0 [11.5, 20.1] 2.89 [1.03, 5.17] 5.61× [2.95, 16.8] 0.0022
Key during motion, VNC No delay Gap between frames, 95th percentile (ms) 70.3 [61.9, 83.6] 50.6 [45.9, 60.5] 1.40× [1.22, 1.64] 0.0022
Key during motion, VNC No delay Data per frame (KiB) 50.4 [49.8, 51.9] 8.02 [7.97, 8.10] 6.27× [6.21, 6.45] 0.0022
Key during motion, VNC No delay Server CPU per frame (ms) 49.3 [45.8, 54.3] 12.9 [11.5, 13.3] 3.89× [3.50, 4.32] 0.0022
Key during motion, VNC No delay Server CPU time (ms) 12,471 [11,923, 13,194] 3,706 [3,407, 3,810] 3.39× [3.24, 3.60] 0.0022
Key during motion, VNC No delay Server memory, peak (MiB) 56.5 [54.7, 57.3] 39.3 [36.4, 41.5] 1.44× [1.36, 1.50] 0.0022
Key during motion, VNC No delay Client CPU time (ms) 4,908 [4,644, 5,302] 2,159 [2,095, 2,242] 2.26× [2.16, 2.38] 0.0022
Key during motion, RDP No delay Frames shown a second (frames/s) 23.3 [21.5, 23.7] 24.3 [24.1, 24.6] 1.04× [1.03, 1.12] 0.0022
Key during motion, RDP No delay Scene frames never shown (%) 22.2 [20.9, 28.2] 19.0 [18.1, 19.8] 1.17× [1.10, 1.42] 0.0022
Key during motion, RDP No delay Gap between frames, 95th percentile (ms) 74.2 [52.9, 89.4] 50.8 [50.0, 57.1] 1.38× [1.06, 1.68] 0.0043
Key during motion, RDP No delay Data per frame (KiB) 66.0 [64.9, 66.7] 7.73 [7.71, 7.84] 8.53× [8.41, 8.63] 0.0022
Key during motion, RDP No delay Server CPU per frame (ms) 58.9 [55.3, 71.2] 15.8 [14.2, 20.2] 3.83× [3.31, 4.48] 0.0022
Key during motion, RDP No delay Server CPU time (ms) 13,938 [13,059, 15,869] 3,855 [3,414, 4,855] 3.70× [3.20, 4.13] 0.0022
Key during motion, RDP No delay Server memory, peak (MiB) 77.9 [67.3, 90.1] 68.4 [64.9, 74.2] 1.15× [1.04, 1.31] 0.0411
Key during motion, RDP No delay Client CPU time (ms) 5,808 [5,334, 6,410] 1,897 [1,680, 2,138] 3.13× [2.74, 3.49] 0.0022
Full-screen photo, VNC No delay Time to complete (ms) 186 [176, 218] 86.8 [73.4, 135] 2.23× [1.61, 2.51] 0.0022
Full-screen photo, VNC No delay Server CPU time (ms) 328 [200, 371] 102 [92.9, 119] 3.23× [2.15, 3.55] 0.0022
Full-screen photo, VNC No delay Server memory, peak (MiB) 61.5 [37.2, 63.0] 41.8 [38.4, 46.6] 1.46× [0.968, 1.57] 0.0649
Full-screen photo, VNC No delay Client CPU time (ms) 85.3 [81.2, 94.2] 51.2 [47.3, 55.2] 1.68× [1.57, 1.82] 0.0022
Full-screen photo, RDP No delay Time to complete (ms) 177 [131, 266] 158 [79.4, 188] 1.39× [0.815, 2.23] 0.3939
Full-screen photo, RDP No delay Server CPU time (ms) 296 [239, 326] 139 [117, 163] 2.06× [1.76, 2.40] 0.0022
Full-screen photo, RDP No delay Server memory, peak (MiB) 70.4 [57.8, 70.9] 67.5 [62.6, 71.4] 1.03× [0.924, 1.09] 0.3095
Full-screen photo, RDP No delay Client CPU time (ms) 93.8 [89.4, 105] 67.5 [53.5, 71.6] 1.47× [1.29, 1.69] 0.0022
Typing, SSH 50 ms, 20 Mbit/s Latency, median (ms) 54.1 [53.6, 55.0] 53.5 [53.0, 54.2] 1.01× [0.994, 1.03] 0.0931
Typing, SSH 50 ms, 20 Mbit/s Server memory, peak (MiB) 47.3 [46.7, 78.5] 56.6 [11.3, 63.5] 0.845× [0.749, 4.13] 0.3095
Typing, SSH 50 ms, 20 Mbit/s Client CPU time (ms) 79.5 [63.6, 95.5] 80.4 [61.4, 97.1] 1.00× [0.76, 1.36] 1.0000
Long output, SSH 50 ms, 20 Mbit/s Data to the client (KiB) 82.7 [59.6, 148] 0.849 [0.699, 1.04] 96.1× [66.3, 182] 0.0022
Long output, SSH 50 ms, 20 Mbit/s Server CPU time (ms) 870 [836, 1,216] 17.7 [14.3, 21.2] 52.5× [43.5, 66.9] 0.0022
Long output, SSH 50 ms, 20 Mbit/s Server memory, peak (MiB) 67.1 [66.4, 68.0] 56.7 [12.4, 64.8] 1.18× [1.04, 5.37] 0.0022
Long output, SSH 50 ms, 20 Mbit/s Client CPU time (ms) 45.2 [35.2, 75.4] 16.5 [12.7, 20.7] 2.83× [1.94, 4.27] 0.0022
Key press, VNC 50 ms, 20 Mbit/s Latency, median (ms) 67.3 [66.3, 67.6] 55.9 [55.2, 56.7] 1.20× [1.18, 1.22] 0.0022
Key press, VNC 50 ms, 20 Mbit/s Server memory, peak (MiB) 75.3 [74.5, 76.7] 61.4 [18.9, 70.5] 1.23× [1.09, 3.95] 0.0022
Key press, VNC 50 ms, 20 Mbit/s Client CPU time (ms) 86.4 [69.0, 109] 105 [74.9, 123] 0.839× [0.666, 1.11] 0.2403
Key press, RDP 50 ms, 20 Mbit/s Latency, median (ms) 69.5 [68.7, 70.0] 58.3 [57.9, 59.1] 1.19× [1.18, 1.20] 0.0022
Key press, RDP 50 ms, 20 Mbit/s Server memory, peak (MiB) 92.5 [67.0, 92.6] 46.7 [34.1, 57.2] 1.88× [1.61, 2.15] 0.0022
Key press, RDP 50 ms, 20 Mbit/s Client CPU time (ms) 82.7 [76.8, 104] 117 [93.6, 142] 0.761× [0.593, 0.918] 0.0260
Key during motion, VNC 50 ms, 20 Mbit/s Frames shown a second (frames/s) 25.3 [24.5, 26.0] 29.2 [28.5, 29.6] 1.15× [1.12, 1.19] 0.0022
Key during motion, VNC 50 ms, 20 Mbit/s Scene frames never shown (%) 15.8 [13.5, 18.4] 2.68 [1.19, 5.04] 5.95× [3.65, 11.3] 0.0022
Key during motion, VNC 50 ms, 20 Mbit/s Gap between frames, 95th percentile (ms) 69.0 [64.6, 72.0] 49.8 [47.5, 51.5] 1.37× [1.30, 1.45] 0.0022
Key during motion, VNC 50 ms, 20 Mbit/s Data per frame (KiB) 46.3 [44.9, 46.9] 7.90 [7.87, 7.97] 5.86× [5.70, 5.93] 0.0022
Key during motion, VNC 50 ms, 20 Mbit/s Server CPU per frame (ms) 46.7 [44.7, 49.2] 12.4 [11.3, 13.8] 3.77× [3.32, 4.10] 0.0022
Key during motion, VNC 50 ms, 20 Mbit/s Server CPU time (ms) 17,164 [16,097, 18,395] 3,603 [3,353, 3,999] 4.76× [4.23, 5.24] 0.0022
Key during motion, VNC 50 ms, 20 Mbit/s Server memory, peak (MiB) 56.3 [55.6, 57.1] 60.2 [38.2, 70.9] 0.936× [0.806, 1.47] 0.3939
Key during motion, VNC 50 ms, 20 Mbit/s Client CPU time (ms) 6,644 [6,179, 7,121] 2,206 [2,041, 2,355] 3.01× [2.75, 3.26] 0.0022
Key during motion, RDP 50 ms, 20 Mbit/s Frames shown a second (frames/s) 23.3 [22.4, 23.7] 24.0 [23.7, 24.7] 1.04× [1.01, 1.06] 0.0022
Key during motion, RDP 50 ms, 20 Mbit/s Scene frames never shown (%) 22.4 [21.0, 25.2] 20.0 [17.6, 20.8] 1.14× [1.04, 1.27] 0.0022
Key during motion, RDP 50 ms, 20 Mbit/s Gap between frames, 95th percentile (ms) 71.4 [60.0, 84.2] 55.4 [44.5, 65.3] 1.30× [1.09, 1.53] 0.0043
Key during motion, RDP 50 ms, 20 Mbit/s Data per frame (KiB) 53.7 [48.8, 56.1] 7.79 [7.64, 7.87] 6.93× [6.39, 7.17] 0.0022
Key during motion, RDP 50 ms, 20 Mbit/s Server CPU per frame (ms) 54.3 [52.1, 57.5] 15.3 [14.3, 18.5] 3.62× [3.10, 3.86] 0.0022
Key during motion, RDP 50 ms, 20 Mbit/s Server CPU time (ms) 19,641 [18,438, 20,991] 3,668 [3,422, 4,428] 5.28× [4.74, 5.78] 0.0022
Key during motion, RDP 50 ms, 20 Mbit/s Server memory, peak (MiB) 73.1 [67.9, 80.4] 65.3 [63.2, 71.8] 1.09× [1.03, 1.23] 0.0152
Key during motion, RDP 50 ms, 20 Mbit/s Client CPU time (ms) 8,207 [7,608, 8,758] 1,854 [1,713, 2,147] 4.33× [3.98, 4.73] 0.0022
Full-screen photo, VNC 50 ms, 20 Mbit/s Time to complete (ms) 1,064 [702, 1,127] 208 [160, 230] 5.09× [4.38, 5.76] 0.0022
Full-screen photo, VNC 50 ms, 20 Mbit/s Server CPU time (ms) 320 [184, 334] 121 [107, 136] 2.54× [1.72, 2.91] 0.0022
Full-screen photo, VNC 50 ms, 20 Mbit/s Server memory, peak (MiB) 61.8 [36.6, 62.8] 46.0 [42.6, 51.3] 1.32× [0.86, 1.44] 0.0649
Full-screen photo, VNC 50 ms, 20 Mbit/s Client CPU time (ms) 87.3 [84.7, 93.3] 50.4 [41.9, 64.2] 1.74× [1.45, 2.02] 0.0022
Full-screen photo, RDP 50 ms, 20 Mbit/s Time to complete (ms) 991 [606, 1,044] 151 [145, 164] 6.37× [4.17, 6.95] 0.0022
Full-screen photo, RDP 50 ms, 20 Mbit/s Server CPU time (ms) 289 [183, 423] 124 [121, 145] 2.31× [1.49, 2.92] 0.0022
Full-screen photo, RDP 50 ms, 20 Mbit/s Server memory, peak (MiB) 62.0 [61.6, 62.5] 64.3 [62.9, 68.7] 0.964× [0.91, 0.983] 0.0022
Full-screen photo, RDP 50 ms, 20 Mbit/s Client CPU time (ms) 89.6 [69.3, 116] 56.7 [49.1, 61.6] 1.63× [1.41, 1.90] 0.0022

The 50 ms target

A key's effect should show within 50 ms on the network with no delay and on the one with a 20 ms round trip. The target is met where the upper bound of the 95% interval of the latency's 95th percentile is at most 50 ms. With no delay, the latency is the server's own processing; the other network adds its round trip.

Series Scenario Network Latency, 95th percentile (ms) Met
Occamy Typing, SSH No delay 1.73 [1.41, 2.63] met
Occamy Key press, VNC No delay 4.64 [4.28, 5.31] met
Occamy Key press, RDP No delay 8.10 [6.92, 12.3] met
Occamy Key during motion, VNC No delay 49.0 [39.6, 54.3] no
Occamy Key during motion, RDP No delay 66.8 [57.9, 92.3] no

occamyd's runtime

For each scenario and network, the median of the runs: the memory occamyd allocated and the garbage collection cycles it ran during the scenario, and its heap in use and its goroutines at the end.

Series Scenario Network Allocated (MiB) GC cycles Heap (MiB) Goroutines
Occamy Typing, SSH No delay 1.31 0 26.1 8.00
Occamy Long output, SSH No delay 4.28 1.50 8.45 8.00
Occamy Key press, VNC No delay 9.59 2.00 12.2 8.00
Occamy Key press, RDP No delay 22.6 2.00 24.5 8.00
Occamy Key during motion, VNC No delay 159 13.0 27.3 8.00
Occamy Key during motion, RDP No delay 87.7 5.00 37.8 8.00
Occamy Full-screen photo, VNC No delay 35.9 3.00 30.3 8.00
Occamy Full-screen photo, RDP No delay 57.4 2.00 51.2 8.00
Occamy Typing, SSH 50 ms, 20 Mbit/s 1.30 0 45.5 8.50
Occamy Long output, SSH 50 ms, 20 Mbit/s 1.68 0 47.5 8.50
Occamy Key press, VNC 50 ms, 20 Mbit/s 10.6 0.5 37.8 8.50
Occamy Key press, RDP 50 ms, 20 Mbit/s 22.4 2.00 25.4 8.00
Occamy Key during motion, VNC 50 ms, 20 Mbit/s 106 9.00 25.8 8.50
Occamy Key during motion, RDP 50 ms, 20 Mbit/s 93.2 5.00 53.3 8.00
Occamy Full-screen photo, VNC 50 ms, 20 Mbit/s 40.9 3.00 35.7 8.50
Occamy Full-screen photo, RDP 50 ms, 20 Mbit/s 54.5 3.50 46.7 8.00

Sensitivity to the host's load

None of the 192 results started while the host's 1-minute load average was above the gate of 8, so no conclusion on a primary metric depends on the host's load.

Figures

Each figure shows a metric's median for each series in each scenario, with its 95% interval, in a panel for each network. Where the values span more than a factor of ten, the axis is logarithmic.

Latency, 95th percentile

Latency, 95th percentile: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkLatency, 95th percentile: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Latency, median

Latency, median: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkLatency, median: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Time to complete

Time to complete: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkTime to complete: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Data to the client

Data to the client: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkData to the client: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Frames shown a second

Frames shown a second: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkFrames shown a second: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Scene frames never shown

Scene frames never shown: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkScene frames never shown: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Gap between frames, 95th percentile

Gap between frames, 95th percentile: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkGap between frames, 95th percentile: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Data per frame

Data per frame: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkData per frame: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Server CPU per frame

Server CPU per frame: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkServer CPU per frame: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Server CPU time

Server CPU time: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkServer CPU time: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Server memory, peak

Server memory, peak: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkServer memory, peak: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Client CPU time

Client CPU time: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkClient CPU time: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

Memory occamyd allocated

Memory occamyd allocated: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkMemory occamyd allocated: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

occamyd's garbage collection cycles

occamyd's garbage collection cycles: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkoccamyd's garbage collection cycles: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

occamyd's heap in use

occamyd's heap in use: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkoccamyd's heap in use: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network

occamyd's goroutines

occamyd's goroutines: each series' median of the runs in each scenario, with its 95% interval, in a panel for each networkoccamyd's goroutines: each series' median of the runs in each scenario, with its 95% interval, in a panel for each network