Add persistent storage to agent

.github/workflows/kind.yml

@@ -19,6 +19,9 @@ jobs:
       - name: Checkout the code
         uses: actions/checkout@v2
 
+      - name: Install qemu-img
+        run: sudo apt update && sudo apt install -y qemu-utils
+
       - name: Set env variables
         run: |
           export "REGISTRY_IP=$(ip addr show eth0 | grep -oP '(?<=inet\s)\d+\.\d+\.\d+\.\d+')"
@@ -73,11 +76,13 @@ jobs:
 
       - name: Deploy assisted-migration
         run: |
-          make deploy-on-kind MIGRATION_PLANNER_API_IMAGE=$MIGRATION_PLANNER_API_IMAGE MIGRATION_PLANNER_AGENT_IMAGE=$MIGRATION_PLANNER_AGENT_IMAGE MIGRATION_PLANNER_API_IMAGE_PULL_POLICY=$MIGRATION_PLANNER_API_IMAGE_PULL_POLICY INSECURE_REGISTRY=$INSECURE_REGISTRY MIGRATION_PLANNER_NAMESPACE=default
+          make deploy-on-kind MIGRATION_PLANNER_API_IMAGE=$MIGRATION_PLANNER_API_IMAGE MIGRATION_PLANNER_AGENT_IMAGE=$MIGRATION_PLANNER_AGENT_IMAGE MIGRATION_PLANNER_API_IMAGE_PULL_POLICY=$MIGRATION_PLANNER_API_IMAGE_PULL_POLICY INSECURE_REGISTRY=$INSECURE_REGISTRY MIGRATION_PLANNER_NAMESPACE=default PERSISTENT_DISK_DEVICE=/dev/vda
           kubectl wait --for=condition=Ready pods --all --timeout=240s
           kubectl port-forward --address 0.0.0.0 service/migration-planner-agent 7443:7443 &
           kubectl port-forward --address 0.0.0.0 service/migration-planner 3443:3443 &
 
       - name: Run test
         run: |
+          mkdir /tmp/untarova
+          qemu-img convert -f vmdk -O qcow2 data/persistence-disk.vmdk /tmp/untarova/persistence-disk.qcow2
           sudo make integration-test PLANNER_IP=${REGISTRY_IP}

Makefile

@@ -144,7 +144,8 @@ deploy-on-kind:
 	sed "s|@MIGRATION_PLANNER_AGENT_IMAGE@|$(MIGRATION_PLANNER_AGENT_IMAGE)|g; \
              s|@INSECURE_REGISTRY@|$(INSECURE_REGISTRY)|g; \
              s|@MIGRATION_PLANNER_API_IMAGE_PULL_POLICY@|$(MIGRATION_PLANNER_API_IMAGE_PULL_POLICY)|g; \
-             s|@MIGRATION_PLANNER_API_IMAGE@|$(MIGRATION_PLANNER_API_IMAGE)|g" \
+             s|@MIGRATION_PLANNER_API_IMAGE@|$(MIGRATION_PLANNER_API_IMAGE)|g; \
+             s|@PERSISTENT_DISK_DEVICE@|$(PERSISTENT_DISK_DEVICE)|g" \
              deploy/k8s/migration-planner.yaml.template > deploy/k8s/migration-planner.yaml
 	$(KUBECTL) apply -n "${MIGRATION_PLANNER_NAMESPACE}" -f 'deploy/k8s/*-service.yaml'
 	$(KUBECTL) apply -n "${MIGRATION_PLANNER_NAMESPACE}" -f 'deploy/k8s/*-secret.yaml'

cmd/planner-agent/main.go

@@ -3,7 +3,6 @@ package main
 import (
 	"bytes"
 	"context"
-	"errors"
 	"flag"
 	"fmt"
 	"os"
@@ -77,15 +76,15 @@ func (a *agentCmd) Execute() error {
 	undo := zap.ReplaceGlobals(logger)
 	defer undo()
 
-	agentID, err := a.readFile(agentFilename)
+	agentID, err := a.readFileFromPersistent(agentFilename)
 	if err != nil {
 		zap.S().Fatalf("failed to retreive agent_id: %v", err)
 	}
 
 	// Try to read jwt from file.
 	// We're assuming the jwt is valid.
 	// The agent will not try to validate the jwt. The backend is responsible for validating the token.
-	jwt, err := a.readFile(jwtFilename)
+	jwt, err := a.readFileFromVolatile(jwtFilename)
 	if err != nil {
 		zap.S().Errorf("failed to read jwt: %v", err)
 	}
@@ -97,16 +96,23 @@ func (a *agentCmd) Execute() error {
 	return nil
 }
 
-func (a *agentCmd) readFile(filename string) (string, error) {
-	// look for it in data dir
-	confDirPath := path.Join(a.config.DataDir, filename)
-	if _, err := os.Stat(confDirPath); err == nil {
-		content, err := os.ReadFile(confDirPath)
+func (a *agentCmd) readFile(baseDir string, filename string) (string, error) {
+	filePath := path.Join(baseDir, filename)
+	if _, err := os.Stat(filePath); err == nil {
+		content, err := os.ReadFile(filePath)
 		if err != nil {
 			return "", err
 		}
 		return string(bytes.TrimSpace(content)), nil
 	}
 
-	return "", errors.New("agent_id not found")
+	return "", fmt.Errorf("file not found: %s", filePath)
+}
+
+func (a *agentCmd) readFileFromVolatile(filename string) (string, error) {
+	return a.readFile(a.config.DataDir, filename)
+}
+
+func (a *agentCmd) readFileFromPersistent(filename string) (string, error) {
+	return a.readFile(a.config.PersistentDataDir, filename)
 }

data/MigrationAssessment.ovf

@@ -1,8 +1,13 @@
 <?xml version='1.0' encoding='UTF-8'?>
 <Envelope xmlns="http://schemas.dmtf.org/ovf/envelope/1" xmlns:ovf="http://schemas.dmtf.org/ovf/envelope/1" xmlns:vmw="http://www.vmware.com/schema/ovf" xmlns:rasd="http://schemas.dmtf.org/wbem/wscim/1/cim-schema/2/CIM_ResourceAllocationSettingData" xmlns:vssd="http://schemas.dmtf.org/wbem/wscim/1/cim-schema/2/CIM_VirtualSystemSettingData">
   <References>
-    <File ovf:id="file1" ovf:href="MigrationAssessment.iso" ovf:size="1220542464"/>
+    <File ovf:href="MigrationAssessment.iso" ovf:id="file1" ovf:size="1244659712"/>
+    <File ovf:href="persistence-disk.vmdk" ovf:id="file2" ovf:size="68096"/>
   </References>
+  <DiskSection>
+    <Info>Virtual disk information</Info>
+    <Disk ovf:capacity="50" ovf:capacityAllocationUnits="byte * 2^20" ovf:diskId="vmdisk1" ovf:fileRef="file2" ovf:format="http://www.vmware.com/interfaces/specifications/vmdk.html#streamOptimized" ovf:populatedSize="0"/>
+  </DiskSection>
   <NetworkSection>
     <Info>The list of logical networks</Info>
     <Network ovf:name="routable-network">
@@ -12,15 +17,16 @@
   <VirtualSystem ovf:id="MigrationAssessment">
     <Info>A Virtual system</Info>
     <Name>MigrationAssessment</Name>
-    <OperatingSystemSection ovf:id="80" ovf:version="8" vmw:osType="rhel9_64Guest">
-      <Info>The operating system installed</Info>
-      <Description>Other Linux (64-bit)</Description>
+    <OperatingSystemSection ovf:id="80" ovf:version="9" vmw:osType="rhel9_64Guest">
+      <Info>The kind of installed guest operating system</Info>
+      <Description>Red Hat Enterprise Linux 9 (64-bit)</Description>
     </OperatingSystemSection>
     <VirtualHardwareSection>
       <Info>Virtual hardware requirements</Info>
       <System>
         <vssd:ElementName>Virtual Hardware Family</vssd:ElementName>
         <vssd:InstanceID>0</vssd:InstanceID>
+        <vssd:VirtualSystemIdentifier>MigrationAssessment</vssd:VirtualSystemIdentifier>
         <vssd:VirtualSystemType>vmx-20</vssd:VirtualSystemType>
       </System>
       <Item>
@@ -44,11 +50,12 @@
         <rasd:Description>SCSI Controller</rasd:Description>
         <rasd:ElementName>SCSI Controller 1</rasd:ElementName>
         <rasd:InstanceID>3</rasd:InstanceID>
+        <rasd:ResourceSubType>VirtualSCSI</rasd:ResourceSubType>
         <rasd:ResourceType>6</rasd:ResourceType>
         <vmw:Config ovf:required="false" vmw:key="slotInfo.pciSlotNumber" vmw:value="16"/>
       </Item>
       <Item>
-        <rasd:Address>0</rasd:Address>
+        <rasd:Address>1</rasd:Address>
         <rasd:Description>IDE Controller</rasd:Description>
         <rasd:ElementName>IDE Controller 1</rasd:ElementName>
         <rasd:InstanceID>4</rasd:InstanceID>
@@ -76,6 +83,17 @@
         <vmw:Config ovf:required="false" vmw:key="wakeOnLanEnabled" vmw:value="true"/>
         <vmw:Config ovf:required="false" vmw:key="connectable.allowGuestControl" vmw:value="false"/>
       </Item>
+      <Item>
+        <rasd:AddressOnParent>0</rasd:AddressOnParent>
+        <rasd:ElementName>Hard disk 1</rasd:ElementName>
+        <rasd:HostResource>ovf:/disk/vmdisk1</rasd:HostResource>
+        <rasd:InstanceID>10</rasd:InstanceID>
+        <rasd:Parent>3</rasd:Parent>
+        <rasd:ResourceType>17</rasd:ResourceType>
+        <vmw:Config ovf:required="false" vmw:key="backing.writeThrough" vmw:value="false"/>
+        <vmw:Config ovf:required="false" vmw:key="backing.diskMode" vmw:value="independent_persistent"/>
+        <vmw:Config ovf:required="false" vmw:key="guestReadOnly" vmw:value="false"/>
+      </Item>
       <Item ovf:required="false">
         <rasd:ElementName>Video card</rasd:ElementName>
         <rasd:InstanceID>7</rasd:InstanceID>

data/ignition.template

@@ -16,15 +16,32 @@ systemd:
       contents: |
         [Unit]
         Description=Service to retrieve system uuid
+        Requires=var-lib-data.mount
+        After=var-lib-data.mount
         [Service]
         Type=oneshot
-        ExecStart=/bin/bash -c 'cat /sys/class/dmi/id/product_uuid > /var/home/core/.migration-planner/data/agent_id'
-        ExecStartPost=/bin/bash -c 'chown core:core /var/home/core/.migration-planner/data/agent_id'
+        ExecStart=/bin/bash -c 'cat /sys/class/dmi/id/product_uuid > /var/lib/data/agent_id'
+        ExecStartPost=/bin/bash -c 'chown -R core:core /var/lib/data'
         RemainAfterExit=true
         [Install]
         WantedBy=multi-user.target
+    - name: var-lib-data.mount
+      enabled: true
+      contents: |
+        [Mount]
+        What=/dev/disk/by-label/DATA
+        Where=/var/lib/data
+        Type=ext4
+        [Install]
+        WantedBy=local-fs.target
 
 storage:
+  filesystems:
+    - path: /var/lib/data
+      device: {{ .PersistentDiskDevice }}
+      format: ext4
+      label: "DATA"
+
   links:
     - path: /home/core/.config/systemd/user/timers.target.wants/podman-auto-update.timer
       target: /usr/lib/systemd/user/podman-auto-update.timer
@@ -88,6 +105,7 @@ storage:
         inline: |
           config-dir: /agent/config
           data-dir: /agent/data
+          persistent-data-dir: /agent/persistent-data
           www-dir: /app/www
           log-level: debug
           update-interval: 5s
@@ -116,14 +134,15 @@ storage:
           Exec= -config /agent/config/config.yaml
           PublishPort=3333:3333
           Volume=/home/core/.migration-planner:/agent:Z
+          Volume=/var/lib/data:/agent/persistent-data:Z
           Environment=OPA_SERVER=127.0.0.1:8181
           Network=host
           UserNS=keep-id:uid=1001
 
           [Service]
           Restart=on-failure
           RestartSec=5
-          ExecStartPre=/bin/bash -c 'until [ -f /var/home/core/.migration-planner/data/agent_id ]; do sleep 1; done'
+          ExecStartPre=/bin/bash -c 'until [ -f /var/lib/data/agent_id ]; do sleep 1; done'
 
           [Install]
           WantedBy=multi-user.target default.target

deploy/k8s/migration-planner.yaml.template

@@ -45,6 +45,8 @@ spec:
               value: @MIGRATION_PLANNER_AGENT_IMAGE@
             - name:  INSECURE_REGISTRY
               value: "@INSECURE_REGISTRY@"
+            - name: PERSISTENT_DISK_DEVICE
+              value: "@PERSISTENT_DISK_DEVICE@"
           volumeMounts:
           volumeMounts:
             - name: migration-planner-config

internal/agent/agent.go

@@ -136,7 +136,7 @@ func (a *Agent) start(ctx context.Context, plannerClient client.Planner) {
 	// TODO refactor health checker to call it from the main goroutine
 	healthChecker.Start(ctx, a.healtCheckStopCh)
 
-	collector := service.NewCollector(a.config.DataDir)
+	collector := service.NewCollector(a.config.DataDir, a.config.PersistentDataDir)
 	collector.Collect(ctx)
 
 	updateTicker := jitterbug.New(time.Duration(a.config.UpdateInterval.Duration), &jitterbug.Norm{Stdev: 30 * time.Millisecond, Mean: 0})

internal/agent/agent_test.go

@@ -68,6 +68,7 @@ var _ = Describe("Agent", func() {
 			config := config.Config{
 				PlannerService:      config.PlannerService{Config: *client.NewDefault()},
 				DataDir:             agentTmpFolder,
+				PersistentDataDir:   agentTmpFolder,
 				ConfigDir:           agentTmpFolder,
 				UpdateInterval:      util.Duration{Duration: updateInterval},
 				HealthCheckInterval: 10,

internal/agent/config/config.go

@@ -26,6 +26,8 @@ const (
 	DefaultConfigFile = DefaultConfigDir + "/config.yaml"
 	// DefaultDataDir is the default directory where the agent's data is stored
 	DefaultDataDir = "/var/lib/planner"
+	// DefaultPersistentDataDir is the default directory where the agent's data is stored
+	DefaultPersistentDataDir = "/var/lib/data"
 	// DefaultWwwDir is the default directory from which the agent serves static files
 	DefaultWwwDir = "/var/www/planner"
 	// DefaultPlannerEndpoint is the default address of the migration planner server
@@ -48,6 +50,8 @@ type Config struct {
 	ConfigDir string `json:"config-dir"`
 	// DataDir is the directory where the agent's data is stored
 	DataDir string `json:"data-dir"`
+	// PersistentDataDir is the directory where persistent data is stored
+	PersistentDataDir string `json:"persistent-data-dir"`
 	// WwwDir is the directory from which the agent serves static files
 	WwwDir string `json:"www-dir"`
 	// SourceID is the ID of this source in the planner
@@ -85,6 +89,7 @@ func NewDefault() *Config {
 	c := &Config{
 		ConfigDir:           DefaultConfigDir,
 		DataDir:             DefaultDataDir,
+		PersistentDataDir:   DefaultPersistentDataDir,
 		WwwDir:              DefaultWwwDir,
 		PlannerService:      PlannerService{Config: *client.NewDefault()},
 		UpdateInterval:      util.Duration{Duration: DefaultUpdateInterval},
@@ -102,14 +107,14 @@ func (cfg *Config) Validate() error {
 	if err := cfg.PlannerService.Validate(); err != nil {
 		return err
 	}
-
 	requiredFields := []struct {
 		value     string
 		name      string
 		checkPath bool
 	}{
 		{cfg.ConfigDir, "config-dir", true},
 		{cfg.DataDir, "data-dir", true},
+		{cfg.PersistentDataDir, "persistent-data-dir", true},
 	}
 
 	for _, field := range requiredFields {

internal/agent/rest.go

@@ -37,7 +37,7 @@ func RegisterApi(router *chi.Mux, statusUpdater *service.StatusUpdater, configur
 		_ = render.Render(w, r, ServiceUIReply{URL: configuration.PlannerService.Service.UI})
 	})
 	router.Put("/api/v1/credentials", func(w http.ResponseWriter, r *http.Request) {
-		credentialHandler(configuration.DataDir, w, r)
+		credentialHandler(configuration.PersistentDataDir, w, r)
 	})
 	router.Get("/api/v1/inventory", func(w http.ResponseWriter, r *http.Request) {
 		data, err := getInventory(configuration.DataDir)

internal/agent/service/collector.go

@@ -29,13 +29,15 @@ import (
 )
 
 type Collector struct {
-	dataDir string
-	once    sync.Once
+	dataDir        string
+	credentialsDir string
+	once           sync.Once
 }
 
-func NewCollector(dataDir string) *Collector {
+func NewCollector(dataDir string, credentialsDir string) *Collector {
 	return &Collector{
-		dataDir: dataDir,
+		dataDir:        dataDir,
+		credentialsDir: credentialsDir,
 	}
 }
 
@@ -56,7 +58,7 @@ func (c *Collector) Collect(ctx context.Context) {
 }
 
 func (c *Collector) run() {
-	credentialsFilePath := filepath.Join(c.dataDir, config.CredentialsFile)
+	credentialsFilePath := filepath.Join(c.credentialsDir, config.CredentialsFile)
 	zap.S().Named("collector").Infof("Waiting for credentials")
 	waitForFile(credentialsFilePath)
 

internal/agent/service/status.go

@@ -54,7 +54,7 @@ func (s *StatusUpdater) UpdateStatus(ctx context.Context, status api.AgentStatus
 
 func (s *StatusUpdater) CalculateStatus() (api.AgentStatus, string, *api.Inventory) {
 	inventoryFilePath := filepath.Join(s.config.DataDir, config.InventoryFile)
-	credentialsFilePath := filepath.Join(s.config.DataDir, config.CredentialsFile)
+	credentialsFilePath := filepath.Join(s.config.PersistentDataDir, config.CredentialsFile)
 	reader := fileio.NewReader()
 
 	err := reader.CheckPathExists(credentialsFilePath)