From 3e90737448170babc3b93e8fe7621729913a71e1 Mon Sep 17 00:00:00 2001
From: rcourtman <courtmanr@gmail.com>
Date: Tue, 11 Nov 2025 22:27:22 +0000
Subject: [PATCH] Fix guest agent OS info calls causing OpenBSD VM crashes
 (related to #692)

Add defensive mitigation to prevent repeated guest-get-osinfo calls that
trigger buggy behavior in QEMU guest agent 9.0.2 on OpenBSD 7.6.

The issue: OpenBSD doesn't have /etc/os-release (Linux convention), and
qemu-ga 9.0.2 appears to spawn excessive helper processes trying to read
this file whenever guest-get-osinfo is called. These helpers don't clean
up properly, eventually exhausting the process table and crashing the VM.

The fix: Track consecutive OS info failures per VM. After 3 failures,
automatically skip future guest-get-osinfo calls for that VM while
continuing to fetch other guest agent data (network interfaces, version).
This prevents triggering the buggy code path while maintaining most guest
agent functionality.

The counter resets on success, so if the guest agent is upgraded or the
issue is resolved, Pulse will automatically resume OS info collection.

Related to #692
---
 internal/monitoring/monitor.go | 80 ++++++++++++++++++++++++----------
 1 file changed, 57 insertions(+), 23 deletions(-)

diff --git a/internal/monitoring/monitor.go b/internal/monitoring/monitor.go
index 22c00c4..04370fa 100644
--- a/internal/monitoring/monitor.go
+++ b/internal/monitoring/monitor.go
@@ -879,15 +879,20 @@ const (
 	defaultGuestAgentVersionTimeout       = 10 * time.Second // GUEST_AGENT_VERSION_TIMEOUT
 	defaultGuestAgentRetries              = 1                // GUEST_AGENT_RETRIES (0 = no retry, 1 = one retry)
 	defaultGuestAgentRetryDelay           = 500 * time.Millisecond
+
+	// Skip OS info calls after this many consecutive failures to avoid triggering buggy guest agents (refs #692)
+	guestAgentOSInfoFailureThreshold = 3
 )
 
 type guestMetadataCacheEntry struct {
-	ipAddresses       []string
-	networkInterfaces []models.GuestNetworkInterface
-	osName            string
-	osVersion         string
-	agentVersion      string
-	fetchedAt         time.Time
+	ipAddresses         []string
+	networkInterfaces   []models.GuestNetworkInterface
+	osName              string
+	osVersion           string
+	agentVersion        string
+	fetchedAt           time.Time
+	osInfoFailureCount  int  // Track consecutive OS info failures
+	osInfoSkip          bool // Skip OS info calls after repeated failures (refs #692)
 }
 
 type taskOutcome struct {
@@ -2464,21 +2469,48 @@ func (m *Monitor) fetchGuestAgentMetadata(ctx context.Context, client PVEClientI
 	}
 
 	// OS info with configurable timeout and retry (refs #592)
-	agentInfoRaw, err := m.retryGuestAgentCall(ctx, m.guestAgentOSInfoTimeout, m.guestAgentRetries, func(ctx context.Context) (interface{}, error) {
-		return client.GetVMAgentInfo(ctx, nodeName, vmid)
-	})
-	if err != nil {
+	// Skip OS info calls if we've seen repeated failures (refs #692 - OpenBSD qemu-ga issue)
+	osInfoFailureCount := cached.osInfoFailureCount
+	osInfoSkip := cached.osInfoSkip
+
+	if !osInfoSkip {
+		agentInfoRaw, err := m.retryGuestAgentCall(ctx, m.guestAgentOSInfoTimeout, m.guestAgentRetries, func(ctx context.Context) (interface{}, error) {
+			return client.GetVMAgentInfo(ctx, nodeName, vmid)
+		})
+		if err != nil {
+			osInfoFailureCount++
+			if osInfoFailureCount >= guestAgentOSInfoFailureThreshold {
+				osInfoSkip = true
+				log.Info().
+					Str("instance", instanceName).
+					Str("vm", vmName).
+					Int("vmid", vmid).
+					Int("failureCount", osInfoFailureCount).
+					Msg("Guest agent OS info consistently fails, skipping future calls to avoid triggering buggy guest agents")
+			} else {
+				log.Debug().
+					Str("instance", instanceName).
+					Str("vm", vmName).
+					Int("vmid", vmid).
+					Int("failureCount", osInfoFailureCount).
+					Err(err).
+					Msg("Guest agent OS info unavailable")
+			}
+		} else if agentInfo, ok := agentInfoRaw.(map[string]interface{}); ok && len(agentInfo) > 0 {
+			osName, osVersion = extractGuestOSInfo(agentInfo)
+			osInfoFailureCount = 0 // Reset on success
+			osInfoSkip = false
+		} else {
+			osName = ""
+			osVersion = ""
+		}
+	} else {
+		// Skipping OS info call due to repeated failures
 		log.Debug().
 			Str("instance", instanceName).
 			Str("vm", vmName).
 			Int("vmid", vmid).
-			Err(err).
-			Msg("Guest agent OS info unavailable")
-	} else if agentInfo, ok := agentInfoRaw.(map[string]interface{}); ok && len(agentInfo) > 0 {
-		osName, osVersion = extractGuestOSInfo(agentInfo)
-	} else {
-		osName = ""
-		osVersion = ""
+			Msg("Skipping guest agent OS info call (disabled after repeated failures)")
 	}
 
 	// Agent version with configurable timeout and retry (refs #592)
@@ -2499,12 +2531,14 @@ func (m *Monitor) fetchGuestAgentMetadata(ctx context.Context, client PVEClientI
 	}
 
 	entry := guestMetadataCacheEntry{
-		ipAddresses:       cloneStringSlice(ipAddresses),
-		networkInterfaces: cloneGuestNetworkInterfaces(networkIfaces),
-		osName:            osName,
-		osVersion:         osVersion,
-		agentVersion:      agentVersion,
-		fetchedAt:         time.Now(),
+		ipAddresses:        cloneStringSlice(ipAddresses),
+		networkInterfaces:  cloneGuestNetworkInterfaces(networkIfaces),
+		osName:             osName,
+		osVersion:          osVersion,
+		agentVersion:       agentVersion,
+		fetchedAt:          time.Now(),
+		osInfoFailureCount: osInfoFailureCount,
+		osInfoSkip:         osInfoSkip,
 	}
 
 	m.guestMetadataMu.Lock()