August was largely a follow-through month: July’s timeout and failure-source model moved from API capability into day-to-day use, while the rollout exposed DNS and timeout edge cases that we tightened across probes and clients.
The Go SDK aligns with the current measurement API
The Go SDK now models measurement status separately from individual probe-result status, supports the current location-selection model, and exposes newer fields such as timeouts and failure sources. This is a breaking update: several API fields that may be null are now pointers, and repeated HTTP headers preserve all returned values instead of collapsing them to one. (globalping-go#13, globalping-go#18)
Table output and timeout controls come to the CLI
Table mode now works across ping, traceroute, MTR, DNS, and HTTP instead of only selected test types, including live result updates. A new global --timeout flag also lets users set a probe-side measurement timeout between 5 and 30 seconds. (globalping-cli#191, globalping-cli#198)
Probe owners can restart probes from the dashboard
Owners of adopted probes can now restart an online probe directly from its dashboard page. After triggering the restart, the dashboard switches to the logs view so the probe’s reconnect can be followed there. (globalping-dash#211, globalping#890, globalping-probe#395)
Fewer false timeouts and clearer failures
July introduced configurable timeouts and failure-source reporting; August focused on edge cases exposed once those features were in use. We raised default timeouts, reserved more of each command’s budget for DNS, retried another configured resolver after resolver failures, fixed false ping and MTR timeouts, and refined failure classification; the website now surfaces target, resolver, and internal errors directly in results. (globalping#892, globalping-probe#402, globalping-probe#398, globalping-probe#397, globalping-probe#411, globalping-probe#403, globalping.io#185)
The blog gets a new look and comments
Globalping’s blog theme was refreshed to match the main site more closely, including navigation, article layouts, responsive behavior, and mobile use. Posts now also support Giscus comments backed by GitHub Discussions, with new posts automatically getting a matching discussion. (ghost-globalping#7, ghost-globalping#8, ghost-globalping#11)
Full Changelog
Below is the complete monthly changelog for Globalping repositories, grouped by project. Maintenance-only changes are omitted.
globalping
- feat: add the restart probe route – globalping#890
- feat: sync probe settings – globalping#893
- fix: increase measurement timeout defaults – globalping#892
- fix: filter for ready probes in scheduled measurements – globalping#895
- fix: use overridden probe locations for limits – globalping#897
globalping.io
- feat: display failure source – globalping.io#185
- feat: add netalis to bronze sponsors – globalping.io#189
- feat: add PipeHost to bronze sponsors – globalping.io#190
- fix: in-progress test issues – globalping.io#187
- fix: probe credits and email – globalping.io#188
- fix: show stopped status for stopped measurements – globalping.io#192
- fix: constrain multi-target table location width – globalping.io#186
- fix: location overflow in single-target table output – globalping.io#193
globalping-probe
- feat: persist probe settings – globalping-probe#405
- fix: gracefully handle probe restart – globalping-probe#395
- fix: retry DNS queries with fallback resolvers – globalping-probe#398
- fix: reserve DNS time in measurement commands – globalping-probe#402
- fix: improve failure source classification – globalping-probe#403
- fix: avoid false ping measurement timeouts – globalping-probe#397
- fix: classify HTTP DNS timeouts as resolver – globalping-probe#396
- fix: classify timeouts after target responses as internal – globalping-probe#401
- fix: handle fractional process timeouts – globalping-probe#411
- perf: increase DNS cache retention – globalping-probe#406
- fix: remove redundant MTR hops after target – globalping-probe#399
- fix: avoid preserving release bundle ownership – globalping-probe#400
- fix: align MTR output columns – globalping-probe#404
globalping-cli
- feat: add table mode to all test types – globalping-cli#191
- feat: add --timeout support – globalping-cli#198
- fix: bind auth callback listener only to localhost – globalping-cli#190
globalping-go
- feat!: align measurement models with the current API schema – globalping-go#13
- fix!: correct nullable and multi-value measurement response fields – globalping-go#18
- feat: add measurement await timeout support – globalping-go#15
- fix: classify measurement await timeouts – globalping-go#16
- fix: send the configured user agent on all requests – globalping-go#17
- fix: preserve HTTP context for invalid error responses – globalping-go#20
globalping-dash
- feat: add the settings tab – globalping-dash#212
- feat: add a probe restart button – globalping-dash#211
- fix: format numbers across the entire app – globalping-dash#213
- fix: keep tag dialog open on prefix change – globalping-dash#216
globalping-dash-directus
globalping-typescript
- fix: limit authentication to secured operations – globalping-typescript#10
- fix: increase default await timeout – globalping-typescript#11
globalping-mcp-server
- fix: correct server domain – globalping-mcp-server#38
- docs: improve setup instructions – globalping-mcp-server#36