fix(collection): correct run status during jobs and add retry-failed

Fix pending mislabeled as fail due to timezone and queue stale checks, show full error details, sync live progress counts, and add retry-failed endpoint for failed NEs only.

Co-authored-by: Cursor <cursoragent@cursor.com>

web/src/i18n/en.ts

@@ -61,6 +61,7 @@ const en = {
     started: "Collection job started: {{id}}",
     paused: "Job paused",
     restarted: "Collection restarted",
+    retryFailedDone: "Retrying failed devices",
     deleted: "Job deleted",
     nothingToRetry: "No failed devices to retry",
     confirmDelete: "Delete this collection job and its log files?",
@@ -80,6 +81,7 @@ const en = {
       collapse: "Hide",
       pause: "Pause",
       restart: "Restart",
+      retryFailed: "Retry failed",
       downloadResults: "Download results",
       delete: "Delete",
       autoRefresh: "Auto-refresh every 2s while jobs are running",

web/src/i18n/zh.ts

@@ -61,6 +61,7 @@ const zh = {
     started: "采集任务已启动：{{id}}",
     paused: "任务已暂停",
     restarted: "已重新开始采集",
+    retryFailedDone: "已开始重采失败网元",
     deleted: "任务已删除",
     nothingToRetry: "没有可重试的失败网元",
     confirmDelete: "确定删除该采集任务？相关日志文件将一并删除。",
@@ -80,6 +81,7 @@ const zh = {
       collapse: "收起",
       pause: "暂停",
       restart: "重新开始",
+      retryFailed: "重采失败网元",
       downloadResults: "下载结果",
       delete: "删除",
       autoRefresh: "任务进行中，每 2 秒自动刷新",