fix(backend): a session or bot that fails to start is retried and the failure is logged

backend/src/watcher/registry.py

@@ -167,17 +167,33 @@ class BotRegistry:
                 await client.stop()
             return
 
-        account_id = await accounts.sync_account(
-            self._pool, me, str(bot_id), is_bot=True
-        )
-        client.capture = await build_bot_capture_context(
-            client, self._pool, self._storage, account_id
-        )
-        await bots.link_account(self._pool, bot_id, account_id)
-        await bots.set_status(
-            self._pool, bot_id, bots.STATUS_OK, username=me.username, title=me.full_name
-        )
-        await accounts.notify_accounts_changed(self._pool)
+        try:
+            account_id = await accounts.sync_account(
+                self._pool, me, str(bot_id), is_bot=True
+            )
+            client.capture = await build_bot_capture_context(
+                client, self._pool, self._storage, account_id
+            )
+            await bots.link_account(self._pool, bot_id, account_id)
+            await bots.set_status(
+                self._pool,
+                bot_id,
+                bots.STATUS_OK,
+                username=me.username,
+                title=me.full_name,
+            )
+            await accounts.notify_accounts_changed(self._pool)
+        except Exception as exc:
+            logger.exception(
+                f"[red]Bot {bot_id} failed to start,[/] retrying on the next sync"
+            )
+            with contextlib.suppress(Exception):
+                await bots.set_status(
+                    self._pool, bot_id, bots.STATUS_ERROR, detail=str(exc)
+                )
+            with contextlib.suppress(Exception):
+                await client.stop()
+            return
 
         consumer = JobConsumer(client, self._pool, account_id)
         self._running[bot_id] = RunningBot(