feat(balloon): add logging for malformed balloon-stats tag

Added warn logging to the decode of BalloonStats if unknown tag is
introduced. Additionally, updated integration test of BalloonStats to
fail if this warn occurs.

The Result type from the update method is removed and corresponding
update metrics have been removed as it is no longer tracked.

Signed-off-by: Aaron Lo <aaronlo0929@gmail.com>

Author: 22aronl

src/vmm/src/devices/virtio/balloon/device.rs

@@ -5,7 +5,7 @@ use std::ops::Deref;
 use std::sync::Arc;
 use std::time::Duration;
 
-use log::{error, info, warn};
+use log::{debug, error, info, warn};
 use serde::{Deserialize, Serialize};
 use utils::time::TimerFd;
 use vmm_sys_util::eventfd::EventFd;
@@ -207,7 +207,7 @@ pub struct BalloonStats {
 }
 
 impl BalloonStats {
-    fn update_with_stat(&mut self, stat: &BalloonStat) -> Result<(), BalloonError> {
+    fn update_with_stat(&mut self, stat: &BalloonStat) {
         let val = Some(stat.val);
         match stat.tag {
             VIRTIO_BALLOON_S_SWAP_IN => self.swap_in = val,
@@ -226,12 +226,8 @@ impl BalloonStats {
             VIRTIO_BALLOON_S_DIRECT_SCAN => self.direct_scan = val,
             VIRTIO_BALLOON_S_ASYNC_RECLAIM => self.async_reclaim = val,
             VIRTIO_BALLOON_S_DIRECT_RECLAIM => self.direct_reclaim = val,
-            _ => {
-                return Err(BalloonError::MalformedPayload);
-            }
+            tag => debug!("balloon: unknown stats update tag: {tag}"),
         }
-
-        Ok(())
     }
 }
 
@@ -503,10 +499,7 @@ impl Balloon {
                 let stat = mem
                     .read_obj::<BalloonStat>(addr)
                     .map_err(|_| BalloonError::MalformedDescriptor)?;
-                self.latest_stats.update_with_stat(&stat).map_err(|_| {
-                    METRICS.stats_update_fails.inc();
-                    BalloonError::MalformedPayload
-                })?;
+                self.latest_stats.update_with_stat(&stat);
             }
 
             self.stats_desc_index = Some(head.index);
@@ -1067,52 +1060,52 @@ pub(crate) mod tests {
             val: 1,
         };
 
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.swap_in, Some(1));
         stat.tag = VIRTIO_BALLOON_S_SWAP_OUT;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.swap_out, Some(1));
         stat.tag = VIRTIO_BALLOON_S_MAJFLT;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.major_faults, Some(1));
         stat.tag = VIRTIO_BALLOON_S_MINFLT;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.minor_faults, Some(1));
         stat.tag = VIRTIO_BALLOON_S_MEMFREE;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.free_memory, Some(1));
         stat.tag = VIRTIO_BALLOON_S_MEMTOT;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.total_memory, Some(1));
         stat.tag = VIRTIO_BALLOON_S_AVAIL;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.available_memory, Some(1));
         stat.tag = VIRTIO_BALLOON_S_CACHES;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.disk_caches, Some(1));
         stat.tag = VIRTIO_BALLOON_S_HTLB_PGALLOC;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.hugetlb_allocations, Some(1));
         stat.tag = VIRTIO_BALLOON_S_HTLB_PGFAIL;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.hugetlb_failures, Some(1));
         stat.tag = VIRTIO_BALLOON_S_OOM_KILL;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.oom_kill, Some(1));
         stat.tag = VIRTIO_BALLOON_S_ALLOC_STALL;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.alloc_stall, Some(1));
         stat.tag = VIRTIO_BALLOON_S_ASYNC_SCAN;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.async_scan, Some(1));
         stat.tag = VIRTIO_BALLOON_S_DIRECT_SCAN;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.direct_scan, Some(1));
         stat.tag = VIRTIO_BALLOON_S_ASYNC_RECLAIM;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.async_reclaim, Some(1));
         stat.tag = VIRTIO_BALLOON_S_DIRECT_RECLAIM;
-        stats.update_with_stat(&stat).unwrap();
+        stats.update_with_stat(&stat);
         assert_eq!(stats.direct_reclaim, Some(1));
     }
 

src/vmm/src/devices/virtio/balloon/metrics.rs

@@ -57,8 +57,6 @@ pub(super) struct BalloonDeviceMetrics {
     pub inflate_count: SharedIncMetric,
     // Number of balloon statistics updates from the driver.
     pub stats_updates_count: SharedIncMetric,
-    // Number of balloon statistics update failures.
-    pub stats_update_fails: SharedIncMetric,
     /// Number of balloon device deflations.
     pub deflate_count: SharedIncMetric,
     /// Number of times when handling events on a balloon device failed.
@@ -83,7 +81,6 @@ impl BalloonDeviceMetrics {
             activate_fails: SharedIncMetric::new(),
             inflate_count: SharedIncMetric::new(),
             stats_updates_count: SharedIncMetric::new(),
-            stats_update_fails: SharedIncMetric::new(),
             deflate_count: SharedIncMetric::new(),
             event_fails: SharedIncMetric::new(),
             free_page_report_count: SharedIncMetric::new(),

src/vmm/src/devices/virtio/block/virtio/metrics.rs

@@ -163,8 +163,6 @@ pub struct BlockDeviceMetrics {
     pub rate_limiter_event_count: SharedIncMetric,
     /// Number of update operation triggered on this block device.
     pub update_count: SharedIncMetric,
-    /// Number of failures while doing update on this block device.
-    pub update_fails: SharedIncMetric,
     /// Number of bytes read by this block device.
     pub read_bytes: SharedIncMetric,
     /// Number of bytes written by this block device.
@@ -215,7 +213,6 @@ impl BlockDeviceMetrics {
         self.rate_limiter_event_count
             .add(other.rate_limiter_event_count.fetch_diff());
         self.update_count.add(other.update_count.fetch_diff());
-        self.update_fails.add(other.update_fails.fetch_diff());
         self.read_bytes.add(other.read_bytes.fetch_diff());
         self.write_bytes.add(other.write_bytes.fetch_diff());
         self.read_count.add(other.read_count.fetch_diff());

tests/integration_tests/functional/test_balloon.py

@@ -320,6 +320,9 @@ def test_stats(uvm_plain_any):
     # Get another reading of the stats after the polling interval has passed.
     deflated_stats = test_microvm.api.balloon_stats.get().json()
 
+    # Ensure that stats don't have unknown balloon stats fields
+    assert "balloon: unknown stats update tag:" not in test_microvm.log_data
+
     # Ensure the stats reflect deflating the balloon.
     assert inflated_stats["free_memory"] < deflated_stats["free_memory"]
     assert inflated_stats["available_memory"] < deflated_stats["available_memory"]
@@ -372,6 +375,9 @@ def test_stats_update(uvm_plain_any):
     final_stats = test_microvm.api.balloon_stats.get().json()
     assert next_stats["available_memory"] != final_stats["available_memory"]
 
+    # Ensure that stats don't have unknown balloon stats fields
+    assert "balloon: unknown stats update tag:" not in test_microvm.log_data
+
 
 def test_balloon_snapshot(uvm_plain_any, microvm_factory):
     """