fix: lambda-promtail, update s3 filename regex to allow finding of log files from AWS GovCloud regions (#12482)

Author: Overflow0xFFFF

tools/lambda-promtail/lambda-promtail/s3.go

@@ -75,9 +75,9 @@ var (
 	// source: https://docs.aws.amazon.com/waf/latest/developerguide/logging-s3.html
 	// format: aws-waf-logs-suffix[/prefix]/AWSLogs/aws-account-id/WAFLogs/region/webacl-name/year/month/day/hour/minute/aws-account-id_waflogs_region_webacl-name_timestamp_hash.log.gz
 	// example: aws-waf-logs-test/AWSLogs/11111111111/WAFLogs/us-east-1/TEST-WEBACL/2021/10/28/19/50/11111111111_waflogs_us-east-1_TEST-WEBACL_20211028T1950Z_e0ca43b5.log.gz
-	defaultFilenameRegex     = regexp.MustCompile(`AWSLogs\/(?P<account_id>\d+)\/(?P<type>[a-zA-Z0-9_\-]+)\/(?P<region>[\w-]+)\/(?P<year>\d+)\/(?P<month>\d+)\/(?P<day>\d+)\/\d+\_(?:elasticloadbalancing|vpcflowlogs)\_\w+-\w+-\d_(?:(?P<lb_type>app|net)\.*?)?(?P<src>[a-zA-Z0-9\-]+)`)
+	defaultFilenameRegex     = regexp.MustCompile(`AWSLogs\/(?P<account_id>\d+)\/(?P<type>[a-zA-Z0-9_\-]+)\/(?P<region>[\w-]+)\/(?P<year>\d+)\/(?P<month>\d+)\/(?P<day>\d+)\/\d+\_(?:elasticloadbalancing|vpcflowlogs)_(?:\w+-\w+-(?:\w+-)?\d)_(?:(?P<lb_type>app|net)\.*?)?(?P<src>[a-zA-Z0-9\-]+)`)
 	defaultTimestampRegex    = regexp.MustCompile(`(?P<timestamp>\d+-\d+-\d+T\d+:\d+:\d+(?:\.\d+Z)?)`)
-	cloudtrailFilenameRegex  = regexp.MustCompile(`AWSLogs\/(?P<organization_id>o-[a-z0-9]{10,32})?\/?(?P<account_id>\d+)\/(?P<type>[a-zA-Z0-9_\-]+)\/(?P<region>[\w-]+)\/(?P<year>\d+)\/(?P<month>\d+)\/(?P<day>\d+)\/\d+\_(?:CloudTrail|CloudTrail-Digest)\_\w+-\w+-\d_(?:(?:app|nlb|net)\.*?)?.+_(?P<src>[a-zA-Z0-9\-]+)`)
+	cloudtrailFilenameRegex  = regexp.MustCompile(`AWSLogs\/(?P<organization_id>o-[a-z0-9]{10,32})?\/?(?P<account_id>\d+)\/(?P<type>[a-zA-Z0-9_\-]+)\/(?P<region>[\w-]+)\/(?P<year>\d+)\/(?P<month>\d+)\/(?P<day>\d+)\/\d+\_(?:CloudTrail|CloudTrail-Digest)_(?:\w+-\w+-(?:\w+-)?\d)_(?:(?:app|nlb|net)\.*?)?.+_(?P<src>[a-zA-Z0-9\-]+)`)
 	cloudfrontFilenameRegex  = regexp.MustCompile(`(?P<prefix>.*)\/(?P<src>[A-Z0-9]+)\.(?P<year>\d+)-(?P<month>\d+)-(?P<day>\d+)-(.+)`)
 	cloudfrontTimestampRegex = regexp.MustCompile(`(?P<timestamp>\d+-\d+-\d+\s\d+:\d+:\d+)`)
 	wafFilenameRegex         = regexp.MustCompile(`AWSLogs\/(?P<account_id>\d+)\/(?P<type>WAFLogs)\/(?P<region>[\w-]+)\/(?P<src>[\w-]+)\/(?P<year>\d+)\/(?P<month>\d+)\/(?P<day>\d+)\/(?P<hour>\d+)\/(?P<minute>\d+)\/\d+\_waflogs\_[\w-]+_[\w-]+_\d+T\d+Z_\w+`)

tools/lambda-promtail/lambda-promtail/s3_test.go

@@ -126,6 +126,39 @@ func Test_getLabels(t *testing.T) {
 			},
 			wantErr: false,
 		},
+		{
+			name: "s3_govcloud_flow_logs",
+			args: args{
+				record: events.S3EventRecord{
+					AWSRegion: "us-gov-east-1",
+					S3: events.S3Entity{
+						Bucket: events.S3Bucket{
+							Name: "vpc_logs_test",
+							OwnerIdentity: events.S3UserIdentity{
+								PrincipalID: "test",
+							},
+						},
+						Object: events.S3Object{
+							Key: "my-bucket/AWSLogs/123456789012/vpcflowlogs/us-gov-east-1/2022/01/24/123456789012_vpcflowlogs_us-gov-east-1_fl-1234abcd_20180620T1620Z_fe123456.log.gz",
+						},
+					},
+				},
+			},
+			want: map[string]string{
+				"account_id":    "123456789012",
+				"bucket":        "vpc_logs_test",
+				"bucket_owner":  "test",
+				"bucket_region": "us-gov-east-1",
+				"day":           "24",
+				"key":           "my-bucket/AWSLogs/123456789012/vpcflowlogs/us-gov-east-1/2022/01/24/123456789012_vpcflowlogs_us-gov-east-1_fl-1234abcd_20180620T1620Z_fe123456.log.gz",
+				"month":         "01",
+				"region":        "us-gov-east-1",
+				"src":           "fl-1234abcd",
+				"type":          FLOW_LOG_TYPE,
+				"year":          "2022",
+			},
+			wantErr: false,
+		},
 		{
 			name: "cloudtrail_digest_logs",
 			args: args{
@@ -192,6 +225,39 @@ func Test_getLabels(t *testing.T) {
 			},
 			wantErr: false,
 		},
+		{
+			name: "cloudtrail_govcloud_logs",
+			args: args{
+				record: events.S3EventRecord{
+					AWSRegion: "us-gov-east-1",
+					S3: events.S3Entity{
+						Bucket: events.S3Bucket{
+							Name: "cloudtrail_logs_test",
+							OwnerIdentity: events.S3UserIdentity{
+								PrincipalID: "test",
+							},
+						},
+						Object: events.S3Object{
+							Key: "my-bucket/AWSLogs/123456789012/CloudTrail/us-gov-east-1/2022/01/24/123456789012_CloudTrail_us-gov-east-1_20220124T0000Z_4jhzXFO2Jlvu2b3y.json.gz",
+						},
+					},
+				},
+			},
+			want: map[string]string{
+				"account_id":    "123456789012",
+				"bucket":        "cloudtrail_logs_test",
+				"bucket_owner":  "test",
+				"bucket_region": "us-gov-east-1",
+				"day":           "24",
+				"key":           "my-bucket/AWSLogs/123456789012/CloudTrail/us-gov-east-1/2022/01/24/123456789012_CloudTrail_us-gov-east-1_20220124T0000Z_4jhzXFO2Jlvu2b3y.json.gz",
+				"month":         "01",
+				"region":        "us-gov-east-1",
+				"src":           "4jhzXFO2Jlvu2b3y",
+				"type":          CLOUDTRAIL_LOG_TYPE,
+				"year":          "2022",
+			},
+			wantErr: false,
+		},
 		{
 			name: "organization_cloudtrail_logs",
 			args: args{
@@ -293,6 +359,41 @@ func Test_getLabels(t *testing.T) {
 			},
 			wantErr: false,
 		},
+		{
+			name: "s3_govcloud_waf",
+			args: args{
+				record: events.S3EventRecord{
+					AWSRegion: "us-gov-east-1",
+					S3: events.S3Entity{
+						Bucket: events.S3Bucket{
+							Name: "waf_logs_test",
+							OwnerIdentity: events.S3UserIdentity{
+								PrincipalID: "test",
+							},
+						},
+						Object: events.S3Object{
+							Key: "prefix/AWSLogs/11111111111/WAFLogs/us-gov-east-1/TEST-WEBACL/2021/10/28/19/50/11111111111_waflogs_us-gov-east-1_TEST-WEBACL_20211028T1950Z_e0ca43b5.log.gz",
+						},
+					},
+				},
+			},
+			want: map[string]string{
+				"account_id":    "11111111111",
+				"bucket_owner":  "test",
+				"bucket_region": "us-gov-east-1",
+				"bucket":        "waf_logs_test",
+				"day":           "28",
+				"hour":          "19",
+				"key":           "prefix/AWSLogs/11111111111/WAFLogs/us-gov-east-1/TEST-WEBACL/2021/10/28/19/50/11111111111_waflogs_us-gov-east-1_TEST-WEBACL_20211028T1950Z_e0ca43b5.log.gz",
+				"minute":        "50",
+				"month":         "10",
+				"region":        "us-gov-east-1",
+				"src":           "TEST-WEBACL",
+				"type":          WAF_LOG_TYPE,
+				"year":          "2021",
+			},
+			wantErr: false,
+		},
 		{
 			name: "missing_type",
 			args: args{