feat: add histogram metric for file access analysis (openobserve#6982)

This histogram metric should allow understanding the age of files
handled by querier.
Example queries
`histogram_quantile(0.95,
rate(zo_file_access_time_bucket{stream_type='logs'}[5m]))`

---------

Co-authored-by: Hengfei Yang <[email protected]>

Author: uddhavdave

src/config/src/metrics.rs

@@ -852,6 +852,21 @@ pub static FILE_DOWNLOADER_PRIORITY_QUEUE_SIZE: Lazy<IntGaugeVec> = Lazy::new(||
     .expect("Metric created")
 });
 
+// File access time bucket histogram
+pub static FILE_ACCESS_TIME: Lazy<HistogramVec> = Lazy::new(|| {
+    HistogramVec::new(
+        HistogramOpts::new(
+            "file_access_time",
+            "Histogram showing query counts within time windows from 1h to 1week (1h, 2h, 3h, 6h, 12h, 24h, 48h, 96h, 168h)"
+        )
+        .namespace(NAMESPACE)
+        .buckets(vec![1.0, 2.0, 3.0, 6.0, 12.0, 24.0, 48.0, 96.0, 168.0])
+        .const_labels(create_const_labels()),
+        &["stream_type"],
+    )
+    .expect("Metric created")
+});
+
 fn register_metrics(registry: &Registry) {
     // http latency
     registry
@@ -1080,6 +1095,9 @@ fn register_metrics(registry: &Registry) {
     registry
         .register(Box::new(FILE_DOWNLOADER_PRIORITY_QUEUE_SIZE.clone()))
         .expect("Metric registered");
+    registry
+        .register(Box::new(FILE_ACCESS_TIME.clone()))
+        .expect("Metric registered");
 }
 
 fn create_const_labels() -> HashMap<String, String> {

src/service/search/grpc/storage.rs

@@ -411,7 +411,8 @@ pub async fn cache_files(
     // check how many files already cached
     let mut cached_files = HashSet::with_capacity(files.len());
     let (mut cache_hits, mut cache_misses) = (0, 0);
-    for (_id, _account, file, _size, _ts) in files.iter() {
+     
+    for (_id, _account, file, _size, max_ts) in files.iter() {
         if file_data::memory::exist(file).await {
             scan_stats.querier_memory_cached_files += 1;
             cached_files.insert(file);
@@ -422,8 +423,35 @@ pub async fn cache_files(
             cache_hits += 1;
         } else {
             cache_misses += 1;
+        };
+         
+        // Record file access metrics
+        let stream_type = if file_type == "index" {
+            config::meta::stream::StreamType::Index
+        } else {
+            // Determine stream type from the file path
+            if file.contains("/logs/") {
+                config::meta::stream::StreamType::Logs
+            } else if file.contains("/metrics/") {
+                config::meta::stream::StreamType::Metrics
+            } else if file.contains("/traces/") {
+                config::meta::stream::StreamType::Traces
+            } else {
+                config::meta::stream::StreamType::Logs // Default
+            }
+        };
+         
+        let current_time = chrono::Utc::now().timestamp_micros();
+        let file_age_seconds = (current_time - max_ts) / 1_000_000;
+        let file_age_hours = file_age_seconds as f64 / 3600.0;
+        
+        if file_age_hours > 0.0  {
+            config::metrics::FILE_ACCESS_TIME
+                .with_label_values(&[&stream_type.to_string()])
+                .observe(file_age_hours);
         }
     }
+
     let files_num = files.len() as i64;
     if files_num == scan_stats.querier_memory_cached_files + scan_stats.querier_disk_cached_files {
         // all files are cached