A Zabbix Client reported that the server entered low memory mode for the Value Cache, even with only ~19% of memory used (4G allocated). The issue was caused by high chunk usage (>90%), as shown in the diaginfo output:

Memory: size: free:3578353424 used:654444480 chunks: free:342450 used:3543114 min:24 max:233392

The customer already monitors zabbix[vcache,cache,free], but since memory was still available, their trigger didn't activate. The root cause was chunk exhaustion, not memory exhaustion — and there is no internal item to monitor chunk usage.

This makes it impossible to proactively detect or trigger alerts on chunk pressure before Value Cache enters degraded mode.

Request: expose chunk metrics (free, used, min, max) via internal items like zabbix[vcache,chunk,...].