 nate
|
cbb1a26582
|
fix: trim chart runs in place and reduce to 25 to match SSR density
|
2026-03-18 20:06:23 +04:00 |
|
nate
|
f2fcf1a0b9
|
fix: evenly space chart runs at fixed width and trim initial data to match
|
2026-03-18 20:02:48 +04:00 |
|
nate
|
766d1094ad
|
fix: evict oldest chart runs as a group instead of individual pings
|
2026-03-18 20:00:11 +04:00 |
|
nate
|
993e6bb1df
|
fix: show full run_id instead of truncating to 8 chars
|
2026-03-18 19:57:41 +04:00 |
|
nate
|
e9ca90324b
|
fix: align chart points by run_id avg time so regions stack vertically
|
2026-03-18 19:55:48 +04:00 |
|
nate
|
1706e83a3f
|
fix: reduce chart spline tension to prevent jittery overshoot
|
2026-03-18 19:52:46 +04:00 |
|
nate
|
59861651bd
|
feat: interactive canvas latency chart with hover tooltips and smooth curves
|
2026-03-18 19:49:44 +04:00 |
|
nate
|
df22029755
|
fix: no region text
|
2026-03-18 19:43:28 +04:00 |
|
nate
|
f013890c40
|
refactor: extract monitor form into shared partial for create and edit
|
2026-03-18 19:41:31 +04:00 |
|
nate
|
5b7a211c21
|
fix: show latest latency from fastest region instead of average on home
|
2026-03-18 19:36:24 +04:00 |
|
nate
|
688245b0c2
|
fix: match client-side sparkline behavior to SSR region-aware rendering
|
2026-03-18 19:31:41 +04:00 |
|
nate
|
6beb7f8039
|
update: tos
|
2026-03-18 19:27:10 +04:00 |
|
nate
|
5be1f85365
|
fix: run cert expiry check concurrently to avoid delaying results
|
2026-03-18 19:04:30 +04:00 |
|
nate
|
42024a9fc3
|
fix: use monitor-side checked_at timestamp instead of coordinator arrival time
|
2026-03-18 18:50:16 +04:00 |
|
nate
|
425bfbfc39
|
perf: optimize monitor runner, fix SSE leak, deduplicate shared utils
|
2026-03-18 18:44:08 +04:00 |
M1
|
980261632e
|
fix: deserialize scheduled_at_ms as string or number (postgres bigint serializes as string)
|
2026-03-18 17:17:29 +04:00 |
|
M1
|
c5eb514990
|
feat: lookahead scheduling — API returns scheduled_at_ms, nodes sleep-until for tight coordination
|
2026-03-18 17:14:28 +04:00 |
|
M1
|
7b98ae78e5
|
fix: run_id = hash(monitor_id, interval_bucket) — unique per window, consistent across regions
|
2026-03-18 17:03:17 +04:00 |
|
M1
|
b6a66ddb21
|
fix: run_id derived from scheduled_at bucket so all regions share same ID per window
|
2026-03-18 16:46:30 +04:00 |
|
M1
|
77534a037c
|
feat: show run_id (first 8 chars) in recent pings table
|
2026-03-18 16:42:50 +04:00 |
|
M1
|
f7ab3b96b2
|
feat: per-region due scheduling + run_id to group pings across regions
|
2026-03-18 16:36:35 +04:00 |
|
M1
|
e057a65535
|
fix: avgLatency NaN — extract .latency_ms from ping objects
|
2026-03-18 16:31:08 +04:00 |
|
M1
|
07648672ad
|
feat: per-region chart lines and lowest-avg sparkline
|
2026-03-18 16:25:47 +04:00 |
|
M1
|
e1bb39431d
|
fix: use URL.searchParams to parse region query param in /internal/due
|
2026-03-18 16:21:24 +04:00 |
|
M1
|
8c39fb45c1
|
feat: add region selector to monitor edit form
|
2026-03-18 16:15:19 +04:00 |
|
M1
|
93db31db3b
|
feat: multi-region monitor support — region selector in UI, region flag on pings
|
2026-03-18 16:08:39 +04:00 |
|
M1
|
52f7f8102b
|
feat: update user-agent to Mozilla-compatible string
|
2026-03-18 14:43:33 +04:00 |
|
M1
|
957674fca4
|
fix: clear in-flight after post_result to prevent double-pings, revert poll to 50ms
|
2026-03-18 14:29:37 +04:00 |
|
M1
|
d3ee0ef9d5
|
fix: revert grace period, increase poll interval to 500ms to prevent doubles without adding jitter
|
2026-03-18 14:27:38 +04:00 |
|
M1
|
5836c44228
|
fix: add 200ms grace period to due query to prevent double-pings
|
2026-03-18 14:21:21 +04:00 |
|
M1
|
6882162d7f
|
fix: use system CA store for ureq TLS to verify Cloudflare-issued certs
|
2026-03-18 14:10:28 +04:00 |
|
M1
|
1b8cbc6e23
|
fix: force ureq to use ring crypto provider to match main.rs init
|
2026-03-18 14:05:04 +04:00 |
|
M1
|
f17bfbfd4b
|
fix: switch ureq from native-tls to rustls (bundled roots)
|
2026-03-18 13:54:39 +04:00 |
|
M1
|
3fa624eff8
|
fix: replace curl subprocess with ureq blocking client in std::thread
|
2026-03-18 13:40:28 +04:00 |
|
M1
|
6b8e1fc9d9
|
cleanup: remove debug warn logs from monitor runner
|
2026-03-18 13:30:19 +04:00 |
|
M1
|
6e1d642c77
|
fix: allow null status_code in ingest schema; check HTTP status in post_result
|
2026-03-18 13:28:42 +04:00 |
|
M1
|
5037222846
|
fix: add 10s timeout to post_result to prevent hung API calls
|
2026-03-18 13:24:52 +04:00 |
|
M1
|
3ce89a88bf
|
debug: log post_result calls
|
2026-03-18 13:22:35 +04:00 |
|
M1
|
749a06b9d7
|
debug: log check errors and trace post_result calls
|
2026-03-18 13:19:35 +04:00 |
|
M1
|
cb8d0f81b1
|
debug: add warn logs to curl thread, fix temp file uniqueness
|
2026-03-18 13:17:11 +04:00 |
|
M1
|
556729b881
|
fix: use std::thread + tokio oneshot for curl — fully decoupled from tokio thread pool
|
2026-03-18 13:10:04 +04:00 |
|
M1
|
5e76b2212f
|
fix: use std::process::Command+spawn_blocking+temp file to avoid all pipe/async hang issues
|
2026-03-18 13:08:08 +04:00 |
|
M1
|
289ec8e038
|
fix: hard task-level timeout as failsafe so in-flight lock always clears
|
2026-03-18 13:05:43 +04:00 |
|
M1
|
0edce8c555
|
fix: use spawn+manual read+wait instead of output() to avoid stdout pipe hang
|
2026-03-18 13:03:06 +04:00 |
|
M1
|
94200b430f
|
fix: set kill_on_drop and explicit stdio pipes so curl doesn't inherit parent fds
|
2026-03-18 12:59:34 +04:00 |
|
M1
|
79bf7d9295
|
fix: wrap curl output() in tokio timeout in case process hangs on stdout
|
2026-03-18 12:58:12 +04:00 |
|
M1
|
554d95cfb4
|
fix: correctly handle curl non-zero exit codes as errors
|
2026-03-18 12:55:00 +04:00 |
|
M1
|
5730a3cb83
|
fix: replace reqwest with curl subprocess for reliable hard timeouts
|
2026-03-18 12:52:17 +04:00 |
|
M1
|
c68700da46
|
fix: use blocking reqwest in spawn_blocking for reliable OS-level timeout
|
2026-03-18 12:48:30 +04:00 |
|
M1
|
4035a3b215
|
fix: switch reqwest to native-tls so connect_timeout covers TLS handshake
|
2026-03-18 12:45:52 +04:00 |