Backbone matrix

How the event backbone (ADR-0070) changes the numbers of the load harness: every scenario on the Native backbone (Phoenix.PubSub + Oban on Postgres) and on Kafka, with one and three nodes. The backbone rows are the backbone zero-loss check: every activity of the run is published to the backbone, and a platform webhook subscriber must observe each one exactly once.

Measured with load/matrix.sh on DESKTOP-Q8TVMOL (2026-10-14).

Commit e1c2862cfe81, 2026-10-14T22:11:16Z, profile full. Generated by load/run.sh (see Benchmarks).

Scenario	Metric	1 node, native	3 nodes, native	1 node, kafka	3 nodes, kafka
e2e	REST→WS p50 (ms)	53	208	41	77
e2e	REST→WS p95 (ms)	114	346	78	144
e2e	REST→WS p99 (ms)	175	394	117	178
e2e	REST POST p95 (ms)	113.3	345.7	77.8	143
e2e	sustained msgs/s (acked)	100	100	100	100
e2e	WS frames/s delivered	1000	1000	1000	1000
e2e	WS connections	1000	1000	1000	1000
webhook_burst	inbound accepted/s	142.8	107.4	128	88
webhook_burst	inbound POST p95 (ms)	1469.5	1803.1	1546.1	1662.5
webhook_burst	inbound→provider p50 (ms)	37272	25373	27623	9826
webhook_burst	inbound→provider p95 (ms)	46747	29070	36564	12653
webhook_burst	inbound→provider p99 (ms)	48297	29716	36985	13389
webhook_burst	deliveries/s	161.6	177.1	164.7	183.1
webhook_burst	queue drained after burst (s)	46.5	20.9	36.6	9.2
idle	idle sockets connected	10000	10000	10000	10000
idle	connect p95 (ms)	6	7	6	7
idle	memory per idle socket (KiB)	108.7	105.8	109.9	113.4
reconnect_storm	reconnect p95 (ms)	887	1061	1545.1	1318
reconnect_storm	reconnect max (ms)	1046	1180	1778	1499
reconnect_storm	REST→WS p95 during storms (ms)	233	461	355	509
rolling_restart	node back to ready, max (s)	10.3	13.7	40.6	43.1
rolling_restart	client reconnect max (ms)	10154	2023	49131	2685
node_kill	killed node back to ready (s)	5.1	7.9	5.7	10
node_kill	client reconnect max (ms)	10080	900	10337	924
backbone	activity events observed / expected	31149	40629	27667	39973
backbone	duplicate events at the subscriber	0	0	0	0
backbone	send→backbone subscriber p50 (ms)	243	2340	357	3823
backbone	send→backbone subscriber p95 (ms)	12484	11493	28162	39689
backbone	send→backbone subscriber p99 (ms)	13609	16783	30227	46839
backbone	caught up after the delivery queues drained (s)	0.3	0.4	0.5	2.2
e2e	lost acked messages	0	0	0	0
webhook_burst	lost acked messages	0	0	0	0
reconnect_storm	lost acked messages	0	0	0	0
rolling_restart	lost acked messages	0	0	0	0
node_kill	lost acked messages	0	0	0	0
node_kill	holes recovered with sync / replay_unavailable errors	54 / 58	325 / 390	25 / 25	420 / 464
backbone	lost acked messages	0	0	0	0

Resources

Highest values sampled while the scenario ran (CPU and RSS about every 2.5 s, the rest every second).

Topology	Scenario	Node	CPU avg %	CPU max %	RSS max (MiB)	DB pool wait max (ms)	Background pool wait max (ms)	Oban queue depth max
1 node, native	e2e	app	391.7	573.7	1428.5	85	15.4	136
1 node, native	e2e	db	151.8	246	304.7	–	–	–
1 node, native	webhook_burst	app	359	581.6	650.2	449.9	16.1	15866
1 node, native	webhook_burst	db	256.8	478.3	367.8	–	–	–
1 node, native	idle	app	59.6	435.3	1627.1	0	340	0
1 node, native	idle	db	14.4	41.8	375	–	–	–
1 node, native	reconnect_storm	app	377.4	595.6	1626.1	98.3	10.1	189
1 node, native	reconnect_storm	db	143.1	265.8	394.7	–	–	–
1 node, native	rolling_restart	app	377.3	589	1610.8	6.8	9.1	297
1 node, native	rolling_restart	db	121.6	213.8	397.5	–	–	–
1 node, native	node_kill	app	346.4	548.7	1380.4	168.4	27.4	671
1 node, native	node_kill	db	116.8	372.6	463.1	–	–	–
3 nodes, native	e2e	app	160.9	216.3	543.9	6.7	27.1	377
3 nodes, native	e2e	app2	150.3	205.5	623.6	6.7	27.1	377
3 nodes, native	e2e	app3	163.8	197.9	581.7	6.7	27.1	377
3 nodes, native	e2e	db	169	241.2	617.6	–	–	–
3 nodes, native	webhook_burst	app	183.5	216.6	456.9	417.1	55.9	11523
3 nodes, native	webhook_burst	app2	183.6	235.4	461.7	417.1	55.9	11523
3 nodes, native	webhook_burst	app3	187.7	241.5	468	417.1	55.9	11523
3 nodes, native	webhook_burst	db	306.2	422.7	707.4	–	–	–
3 nodes, native	idle	app	26	71.4	781.4	0.1	223.3	0
3 nodes, native	idle	app2	26.7	68.8	797.9	0.1	223.3	0
3 nodes, native	idle	app3	25.1	72.6	820.6	0.1	223.3	0
3 nodes, native	idle	db	15.5	45.8	718.2	–	–	–
3 nodes, native	reconnect_storm	app	160.8	232.3	809.3	139.3	35.8	451
3 nodes, native	reconnect_storm	app2	154.5	212.8	827.5	139.3	35.8	451
3 nodes, native	reconnect_storm	app3	160.4	246.5	904.9	139.3	35.8	451
3 nodes, native	reconnect_storm	db	155	291	769.2	–	–	–
3 nodes, native	rolling_restart	app	177.9	347.2	1123.3	168.1	54.3	1061
3 nodes, native	rolling_restart	app2	153.1	286.1	774.1	168.1	54.3	1061
3 nodes, native	rolling_restart	app3	153.8	305.2	1216.5	168.1	54.3	1061
3 nodes, native	rolling_restart	db	164	302.8	767.7	–	–	–
3 nodes, native	node_kill	app	131.1	219.4	754	206.8	19.7	123
3 nodes, native	node_kill	app2	176.1	293.1	734.2	206.8	19.7	123
3 nodes, native	node_kill	app3	152.8	239.3	666.4	206.8	19.7	123
3 nodes, native	node_kill	db	168.2	373.1	851.4	–	–	–
1 node, kafka	e2e	app	438.7	544.3	883.4	5	0.7	20
1 node, kafka	e2e	db	154.3	231.9	292.8	–	–	–
1 node, kafka	e2e	kafka	57.1	214.7	899.4	–	–	–
1 node, kafka	webhook_burst	app	365.6	477.3	622.9	324.1	2.2	10266
1 node, kafka	webhook_burst	db	308	496.9	354	–	–	–
1 node, kafka	webhook_burst	kafka	45.1	177.9	1070.1	–	–	–
1 node, kafka	idle	app	69.2	444.8	1664	0.1	118	0
1 node, kafka	idle	db	14.3	54	359.1	–	–	–
1 node, kafka	idle	kafka	31.5	170.3	1076.2	–	–	–
1 node, kafka	reconnect_storm	app	413.9	572.4	1700.9	106.1	2.7	74
1 node, kafka	reconnect_storm	db	171.7	318.1	390.6	–	–	–
1 node, kafka	reconnect_storm	kafka	46.7	181	1085.4	–	–	–
1 node, kafka	rolling_restart	app	239.6	536.4	1735.7	136.1	1.4	211
1 node, kafka	rolling_restart	db	81.6	237.3	424.1	–	–	–
1 node, kafka	rolling_restart	kafka	35.4	193.1	1093.6	–	–	–
1 node, kafka	node_kill	app	374.3	548	1379.3	95	3.1	97
1 node, kafka	node_kill	db	130.1	331.9	438	–	–	–
1 node, kafka	node_kill	kafka	45.9	197.9	1109	–	–	–
3 nodes, kafka	e2e	app	186.1	222.3	601.4	16.7	1.1	17
3 nodes, kafka	e2e	app2	184.2	221.5	522.1	16.7	1.1	17
3 nodes, kafka	e2e	app3	178.9	221.6	491.2	16.7	1.1	17
3 nodes, kafka	e2e	db	193.5	280.1	618	–	–	–
3 nodes, kafka	e2e	kafka	58.2	179.6	811.2	–	–	–
3 nodes, kafka	webhook_burst	app	194.7	220.6	484.4	375.5	11.2	4000
3 nodes, kafka	webhook_burst	app2	192.3	218.2	469.7	375.5	11.2	4000
3 nodes, kafka	webhook_burst	app3	199.3	240.9	474.4	375.5	11.2	4000
3 nodes, kafka	webhook_burst	db	365.4	439.7	705.1	–	–	–
3 nodes, kafka	webhook_burst	kafka	46.4	140.4	968.3	–	–	–
3 nodes, kafka	idle	app	29.9	193.4	828.1	0.1	9.7	11
3 nodes, kafka	idle	app2	31.3	186.3	800.9	0.1	9.7	11
3 nodes, kafka	idle	app3	31.1	184.9	813	0.1	9.7	11
3 nodes, kafka	idle	db	16	80.6	714.3	–	–	–
3 nodes, kafka	idle	kafka	36.2	187.4	1028.1	–	–	–
3 nodes, kafka	reconnect_storm	app	175.6	235.3	821.1	157	15	27
3 nodes, kafka	reconnect_storm	app2	175.3	228.8	870.7	157	15	27
3 nodes, kafka	reconnect_storm	app3	172.7	236.1	831.2	157	15	27
3 nodes, kafka	reconnect_storm	db	168.7	328.9	763.6	–	–	–
3 nodes, kafka	reconnect_storm	kafka	50.8	182.6	1036.3	–	–	–
3 nodes, kafka	rolling_restart	app	217.2	347.8	1042.4	110.3	2.5	30
3 nodes, kafka	rolling_restart	app2	152.3	279.8	757	110.3	2.5	30
3 nodes, kafka	rolling_restart	app3	174.7	370.4	1202.2	110.3	2.5	30
3 nodes, kafka	rolling_restart	db	175.9	357.8	799.5	–	–	–
3 nodes, kafka	rolling_restart	kafka	47.3	192.9	1061.9	–	–	–
3 nodes, kafka	node_kill	app	154.6	230.1	831.6	176.3	5.9	25
3 nodes, kafka	node_kill	app2	201.2	275.9	812.3	176.3	5.9	25
3 nodes, kafka	node_kill	app3	182	243.1	622	176.3	5.9	25
3 nodes, kafka	node_kill	db	183.9	344.2	838.8	–	–	–
3 nodes, kafka	node_kill	kafka	45.1	196.3	1069.1	–	–	–

Setup

• 1 node, native: 12 vCPU and 30.8 GiB visible to Docker (MINGW64_NT-10.0-26200, local, kernel 6.6.87.2-microsoft-standard-WSL2, Docker 29.8.2); per-node limits CPU unlimited, memory unlimited; Postgres 17.11, 217.1 MiB after the run; pool size 20 + 10 background; mock provider latency 20+10 ms, error rate 0.
  • e2e: conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60
  • webhook_burst: rate=500, duration_s=30, participants=200, targets=3
  • idle: sockets=10000, sockets_per_vu=200, connect_rate=100, ramp_s=100, hold_s=60
  • reconnect_storm: mode=storm, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60, storm_at_s=15, storm_every_s=15
  • rolling_restart: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=70, storm_at_s=null, storm_every_s=null
  • node_kill: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=70, storm_at_s=null, storm_every_s=null
  • backbone: backbone=native, subscriber=platform webhook (activity.created)
• 3 nodes, native: 12 vCPU and 30.8 GiB visible to Docker (MINGW64_NT-10.0-26200, local, kernel 6.6.87.2-microsoft-standard-WSL2, Docker 29.8.2); per-node limits CPU unlimited, memory unlimited; Postgres 17.11, 265.7 MiB after the run; pool size 20 + 10 background; mock provider latency 20+10 ms, error rate 0.
  • e2e: conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60
  • webhook_burst: rate=500, duration_s=30, participants=200, targets=3
  • idle: sockets=10000, sockets_per_vu=200, connect_rate=100, ramp_s=100, hold_s=60
  • reconnect_storm: mode=storm, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60, storm_at_s=15, storm_every_s=15
  • rolling_restart: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=170, storm_at_s=null, storm_every_s=null
  • node_kill: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=70, storm_at_s=null, storm_every_s=null
  • backbone: backbone=native, subscriber=platform webhook (activity.created)
• 1 node, kafka: 12 vCPU and 30.8 GiB visible to Docker (MINGW64_NT-10.0-26200, local, kernel 6.6.87.2-microsoft-standard-WSL2, Docker 29.8.2); per-node limits CPU unlimited, memory unlimited; Postgres 17.11, 197.2 MiB after the run; pool size 20 + 10 background; mock provider latency 20+10 ms, error rate 0.
  • e2e: conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60
  • webhook_burst: rate=500, duration_s=30, participants=200, targets=3
  • idle: sockets=10000, sockets_per_vu=200, connect_rate=100, ramp_s=100, hold_s=60
  • reconnect_storm: mode=storm, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60, storm_at_s=15, storm_every_s=15
  • rolling_restart: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=70, storm_at_s=null, storm_every_s=null
  • node_kill: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=70, storm_at_s=null, storm_every_s=null
  • backbone: backbone=kafka, subscriber=platform webhook (activity.created)
• 3 nodes, kafka: 12 vCPU and 30.8 GiB visible to Docker (MINGW64_NT-10.0-26200, local, kernel 6.6.87.2-microsoft-standard-WSL2, Docker 29.8.2); per-node limits CPU unlimited, memory unlimited; Postgres 17.11, 262.8 MiB after the run; pool size 20 + 10 background; mock provider latency 20+10 ms, error rate 0.
  • e2e: conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60
  • webhook_burst: rate=500, duration_s=30, participants=200, targets=3
  • idle: sockets=10000, sockets_per_vu=200, connect_rate=100, ramp_s=100, hold_s=60
  • reconnect_storm: mode=storm, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=60, storm_at_s=15, storm_every_s=15
  • rolling_restart: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=170, storm_at_s=null, storm_every_s=null
  • node_kill: mode=restart, conversations=100, subscribers_per_conversation=10, sockets=1000, rate_per_conversation=1, duration_s=70, storm_at_s=null, storm_every_s=null
  • backbone: backbone=kafka, subscriber=platform webhook (activity.created)

Checks

• ✅ 1 node, native: e2e: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, native: e2e: messages sent and received live (6000 acked, 60000 latency samples)
• ✅ 1 node, native: webhook_burst: k6 ran and correctness thresholds passed (k6 exit 99)
• ⚠️ 1 node, native: webhook_burst: performance targets (missed: http_req_duration{name:POST /inbound}: p(95)<500, dropped_iterations: count<1, inbound_rejected: count<1)
• ✅ 1 node, native: webhook_burst: every accepted message stored and delivered to every target (accepted 4284, stored 4284, delivered 12852/12852)
• ✅ 1 node, native: idle: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, native: reconnect_storm: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, native: reconnect_storm: messages sent and received live (5996 acked, 59607 latency samples)
• ✅ 1 node, native: rolling_restart: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, native: rolling_restart: messages sent and received live (6740 acked, 64048 latency samples)
• ✅ 1 node, native: node_kill: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, native: node_kill: messages sent and received live (6629 acked, 64623 latency samples)
• ✅ 1 node, native: backbone (native): every activity observed by the subscriber exactly once (31149/31149 observed, 31149 webhook events (0 duplicates), 1 HTTP redeliveries)
• ✅ 1 node, native: e2e: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, native: reconnect_storm: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, native: rolling_restart: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, native: node_kill: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, native: fan-out channel: every activity delivered to the outbound webhook (26865/26865 delivered, 0 jobs still open)
• ✅ 3 nodes, native: e2e: k6 ran and correctness thresholds passed (k6 exit 99)
• ⚠️ 3 nodes, native: e2e: performance targets (missed: e2e_latency_ms: p(95)<250)
• ✅ 3 nodes, native: e2e: messages sent and received live (6000 acked, 60000 latency samples)
• ✅ 3 nodes, native: webhook_burst: k6 ran and correctness thresholds passed (k6 exit 99)
• ⚠️ 3 nodes, native: webhook_burst: performance targets (missed: inbound_rejected: count<1, dropped_iterations: count<1, http_req_duration{name:POST /inbound}: p(95)<500)
• ✅ 3 nodes, native: webhook_burst: every accepted message stored and delivered to every target (accepted 3223, stored 3223, delivered 9669/9669)
• ✅ 3 nodes, native: idle: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, native: reconnect_storm: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, native: reconnect_storm: messages sent and received live (5986 acked, 59002 latency samples)
• ✅ 3 nodes, native: rolling_restart: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, native: rolling_restart: messages sent and received live (16965 acked, 168393 latency samples)
• ✅ 3 nodes, native: node_kill: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, native: node_kill: messages sent and received live (6955 acked, 69378 latency samples)
• ✅ 3 nodes, native: backbone (native): every activity observed by the subscriber exactly once (40629/40629 observed, 40629 webhook events (0 duplicates), 0 HTTP redeliveries)
• ✅ 3 nodes, native: e2e: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, native: reconnect_storm: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, native: rolling_restart: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, native: node_kill: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, native: fan-out channel: every activity delivered to the outbound webhook (37406/37406 delivered, 0 jobs still open)
• ✅ 1 node, kafka: e2e: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, kafka: e2e: messages sent and received live (6000 acked, 60000 latency samples)
• ✅ 1 node, kafka: webhook_burst: k6 ran and correctness thresholds passed (k6 exit 99)
• ⚠️ 1 node, kafka: webhook_burst: performance targets (missed: inbound_rejected: count<1, dropped_iterations: count<1, http_req_duration{name:POST /inbound}: p(95)<500)
• ✅ 1 node, kafka: webhook_burst: every accepted message stored and delivered to every target (accepted 3840, stored 3840, delivered 11520/11520)
• ✅ 1 node, kafka: idle: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, kafka: reconnect_storm: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, kafka: reconnect_storm: messages sent and received live (5996 acked, 59270 latency samples)
• ✅ 1 node, kafka: rolling_restart: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, kafka: rolling_restart: messages sent and received live (3766 acked, 30787 latency samples)
• ✅ 1 node, kafka: node_kill: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 1 node, kafka: node_kill: messages sent and received live (6565 acked, 63995 latency samples)
• ✅ 1 node, kafka: backbone (kafka): every activity observed by the subscriber exactly once (27667/27667 observed, 27667 webhook events (0 duplicates), 3 HTTP redeliveries)
• ✅ 1 node, kafka: e2e: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, kafka: reconnect_storm: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, kafka: rolling_restart: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, kafka: node_kill: zero lost acked messages (lost 0, never acked 0)
• ✅ 1 node, kafka: fan-out channel: every activity delivered to the outbound webhook (23827/23827 delivered, 0 jobs still open)
• ✅ 3 nodes, kafka: e2e: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, kafka: e2e: messages sent and received live (6000 acked, 60000 latency samples)
• ✅ 3 nodes, kafka: webhook_burst: k6 ran and correctness thresholds passed (k6 exit 99)
• ⚠️ 3 nodes, kafka: webhook_burst: performance targets (missed: dropped_iterations: count<1, http_req_duration{name:POST /inbound}: p(95)<500, inbound_rejected: count<1)
• ✅ 3 nodes, kafka: webhook_burst: every accepted message stored and delivered to every target (accepted 2639, stored 2639, delivered 7917/7917)
• ✅ 3 nodes, kafka: idle: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, kafka: reconnect_storm: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, kafka: reconnect_storm: messages sent and received live (5976 acked, 58975 latency samples)
• ✅ 3 nodes, kafka: rolling_restart: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, kafka: rolling_restart: messages sent and received live (16941 acked, 167765 latency samples)
• ✅ 3 nodes, kafka: node_kill: k6 ran and correctness thresholds passed (k6 exit 0)
• ✅ 3 nodes, kafka: node_kill: messages sent and received live (6917 acked, 68964 latency samples)
• ✅ 3 nodes, kafka: backbone (kafka): every activity observed by the subscriber exactly once (39973/39973 observed, 39973 webhook events (0 duplicates), 0 HTTP redeliveries)
• ✅ 3 nodes, kafka: e2e: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, kafka: reconnect_storm: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, kafka: rolling_restart: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, kafka: node_kill: zero lost acked messages (lost 0, never acked 0)
• ✅ 3 nodes, kafka: fan-out channel: every activity delivered to the outbound webhook (37334/37334 delivered, 0 jobs still open)