Merge pull request #1044 from criblio/feat/1025-psm-uuid

1025 Add machine ID and UUID to process start message
criblio · Jul 21, 2022 · 431fb7a · 431fb7a
2 parents 2a692c2 + 7a08f1e
commit 431fb7a
Show file tree

Hide file tree

Showing 13 changed files with 307 additions and 8 deletions.
diff --git a/docs/schemas/definitions/data.schema.json b/docs/schemas/definitions/data.schema.json
@@ -294,6 +294,11 @@
       "type": "string",
       "enum": ["debug", "info", "warning", "error", "none"]
     },
+    "machine_id" : {
+      "title": "machine_id",
+      "description": "A unique identifier for a machine.",
+      "type": "string",
+    },
     "maxeventpersec": {
       "title": "maxeventpersec",
       "description": "Specifies a rate limit for events. See `scope.yml`.",
@@ -625,6 +630,11 @@
       "type": "string",
       "const": "thread"
     },
+    "uuid" : {
+      "title": "uuid",
+      "description": "A unique identifier for a process.",
+      "type": "string",
+    },
     "validateserver": {
       "title": "validateserver",
       "description": "Specifies whether to validate the TLS server certificate. See `scope.yml`.",

diff --git a/docs/schemas/event_start_msg.schema.json b/docs/schemas/event_start_msg.schema.json
@@ -4,7 +4,7 @@
   "type": "object",
   "title": "AppScope Start message",
   "description": "Structure of the start message",
-  "examples": [{"format":"ndjson","info":{"process":{"libscopever":"v1.1.0","pid":35673,"ppid":3390,"gid":1000,"groupname":"test_user","uid":1000,"username":"test_user","hostname":"test_user","procname":"ls","cmd":"ls --color=auto","id":"test_user-ls-ls --color=auto"},"configuration":{"current":{"metric":{"enable":"true","transport":{"type":"udp","host":"127.0.0.1","port":"8125","tls":{"enable":"false","validateserver":"true","cacertpath":""}},"format":{"type":"statsd","statsdprefix":"","statsdmaxlen":512,"verbosity":4},"watch":[{"type":"fs"},{"type":"net"},{"type":"http"},{"type":"dns"},{"type":"process"},{"type":"statsd"}]},"libscope":{"log":{"level":"info","transport":{"type":"file","path":"/tmp/scope.log","buffering":"line"}},"configevent":"true","summaryperiod":10,"commanddir":"/tmp"},"event":{"enable":"true","transport":{"type":"tcp","host":"127.0.0.1","port":"9109","tls":{"enable":"false","validateserver":"true","cacertpath":""}},"format":{"type":"ndjson","maxeventpersec":10000,"enhancefs":"true"},"watch":[{"type":"file","name":"(\\/logs?\\/)|(\\.log$)|(\\.log[.\\d])","field":".*","value":".*"},{"type":"console","name":"(stdout)|(stderr)","field":".*","value":".*","allowbinary":"true"},{"type":"http","name":".*","field":".*","value":".*","headers":[]},{"type":"net","name":".*","field":".*","value":".*"},{"type":"fs","name":".*","field":".*","value":".*"},{"type":"dns","name":".*","field":".*","value":".*"}]},"payload":{"enable":"false","dir":"/tmp"},"tags":{},"protocol":[],"cribl":{"enable":"false","transport":{"type":"edge"},"authtoken":""}}},"environment":{}}}],
+  "examples": [{"format":"ndjson","info":{"process":{"libscopever":"v1.1.0","pid":35673,"ppid":3390,"gid":1000,"groupname":"test_user","uid":1000,"username":"test_user","hostname":"test_user","procname":"ls","cmd":"ls --color=auto","id":"test_user-ls-ls --color=auto","machine_id":"a1e2ada5a5b1b273b4b5c0c2c1c4f5d1","uuid":"da845a9b-a55d-4c42-893d-08b54ee6e999"},"configuration":{"current":{"metric":{"enable":"true","transport":{"type":"udp","host":"127.0.0.1","port":"8125","tls":{"enable":"false","validateserver":"true","cacertpath":""}},"format":{"type":"statsd","statsdprefix":"","statsdmaxlen":512,"verbosity":4},"watch":[{"type":"fs"},{"type":"net"},{"type":"http"},{"type":"dns"},{"type":"process"},{"type":"statsd"}]},"libscope":{"log":{"level":"info","transport":{"type":"file","path":"/tmp/scope.log","buffering":"line"}},"configevent":"true","summaryperiod":10,"commanddir":"/tmp"},"event":{"enable":"true","transport":{"type":"tcp","host":"127.0.0.1","port":"9109","tls":{"enable":"false","validateserver":"true","cacertpath":""}},"format":{"type":"ndjson","maxeventpersec":10000,"enhancefs":"true"},"watch":[{"type":"file","name":"(\\/logs?\\/)|(\\.log$)|(\\.log[.\\d])","field":".*","value":".*"},{"type":"console","name":"(stdout)|(stderr)","field":".*","value":".*","allowbinary":"true"},{"type":"http","name":".*","field":".*","value":".*","headers":[]},{"type":"net","name":".*","field":".*","value":".*"},{"type":"fs","name":".*","field":".*","value":".*"},{"type":"dns","name":".*","field":".*","value":".*"}]},"payload":{"enable":"false","dir":"/tmp"},"tags":{},"protocol":[],"cribl":{"enable":"false","transport":{"type":"edge"},"authtoken":""}}},"environment":{}}}],
   "required": [
     "format",
     "info"
@@ -44,7 +44,9 @@
             "hostname",
             "procname",
             "cmd",
-            "id"
+            "id",
+            "machine_id",
+            "uuid"
           ],
           "properties": {
             "libscopever": {
@@ -79,6 +81,12 @@
             },
             "id": {
               "$ref": "definitions/data.schema.json#/$defs/id"
+            },
+            "machine_id": {
+              "$ref": "definitions/data.schema.json#/$defs/machine_id"
+            },
+            "uuid": {
+              "$ref": "definitions/data.schema.json#/$defs/uuid"
             }
           }
         },

diff --git a/os/linux/Makefile b/os/linux/Makefile
@@ -99,6 +99,7 @@ coretest: $(C_FILES) $(YAML_AR) $(JSON_AR) $(TEST_LIB)
 	$(CC) $(TEST_CFLAGS) -o test/$(OS)/cfgtest cfgtest.o cfg.o scopestdlib.o dbg.o test.o $(TEST_AR) $(TEST_LD_FLAGS)
 	$(CC) $(TEST_CFLAGS) -o test/$(OS)/transporttest transporttest.o transport.o scopestdlib.o dbg.o log.o fn.o utils.o os.o test.o $(TEST_AR) $(TEST_LD_FLAGS)
 	$(CC) $(TEST_CFLAGS) -o test/$(OS)/logtest logtest.o log.o transport.o scopestdlib.o dbg.o fn.o utils.o os.o test.o $(TEST_AR) $(TEST_LD_FLAGS)
+	$(CC) $(TEST_CFLAGS) -o test/$(OS)/utilstest utilstest.o scopestdlib.o dbg.o fn.o utils.o os.o test.o $(TEST_AR) $(TEST_LD_FLAGS)
 	$(CC) $(TEST_CFLAGS) -o test/$(OS)/mtctest mtctest.o mtc.o log.o transport.o mtcformat.o strset.o com.o ctl.o evtformat.o cfg.o cfgutils.o scopestdlib.o dbg.o circbuf.o linklist.o fn.o utils.o os.o test.o report.o search.o httpagg.o state.o httpstate.o metriccapture.o plattime.o $(TEST_AR) $(TEST_LD_FLAGS) -Wl,--wrap=cfgLogStreamEnable
 	$(CC) $(TEST_CFLAGS) -o test/$(OS)/evtformattest evtformattest.o evtformat.o log.o transport.o mtcformat.o strset.o scopestdlib.o dbg.o cfg.o com.o ctl.o mtc.o circbuf.o cfgutils.o linklist.o fn.o utils.o os.o test.o report.o search.o httpagg.o state.o httpstate.o metriccapture.o plattime.o $(TEST_AR) $(TEST_LD_FLAGS)
 	$(CC) $(TEST_CFLAGS) -o test/$(OS)/ctltest ctltest.o ctl.o log.o transport.o scopestdlib.o dbg.o cfgutils.o cfg.o com.o mtc.o evtformat.o mtcformat.o strset.o circbuf.o linklist.o fn.o utils.o os.o test.o report.o search.o httpagg.o state.o httpstate.o metriccapture.o plattime.o $(TEST_AR) $(TEST_LD_FLAGS) -Wl,--wrap=cbufGet

diff --git a/src/com.c b/src/com.c
@@ -173,8 +173,9 @@ jsonProcessObject(proc_id_t *proc)
         if (!(cJSON_AddStringToObjLN(root, "cmd", proc->cmd))) goto err;
     }
     if (!(cJSON_AddStringToObjLN(root, "id", proc->id))) goto err;
-    // starttime
-
+    if (!(cJSON_AddStringToObjLN(root, "machine_id", proc->machine_id))) goto err;
+    if (!(cJSON_AddStringToObjLN(root, "uuid", proc->uuid))) goto err;
+
     return root;
 err:
     if (root) cJSON_Delete(root);

diff --git a/src/scopestdlib.c b/src/scopestdlib.c
@@ -196,6 +196,8 @@ extern void*         scopelibc_shmat(int, const void *, int);
 extern int           scopelibc_shmdt(const void *);
 extern int           scopelibc_shmget(key_t, size_t, int);
 extern int           scopelibc_sched_getcpu(void);
+extern int           scopelibc_rand(void);
+extern void          scopelibc_srand(unsigned int);
 
 static int g_go_static;
 
@@ -1156,3 +1158,16 @@ scope___ctype_tolower_loc(void)
 {
     return scopelibc___ctype_tolower_loc();
 }
+
+int
+scope_rand(void)
+{
+    return scopelibc_rand();
+}
+
+void
+scope_srand(unsigned int seed)
+{
+    scopelibc_srand(seed);
+}
+
diff --git a/src/scopestdlib.h b/src/scopestdlib.h
@@ -250,5 +250,8 @@ void*         scope_shmat(int, const void *, int);
 int           scope_shmdt(const void *);
 int           scope_shmget(key_t, size_t, int);
 int           scope_sched_getcpu(void);
+int           scope_rand(void);
+void          scope_srand(unsigned int);
+
 
 #endif // __SCOPE_STDLIB_H__
diff --git a/src/scopetypes.h b/src/scopetypes.h
@@ -37,6 +37,8 @@ typedef enum {CFG_MTC_FS,
 
 #define MAX_HOSTNAME 255
 #define MAX_PROCNAME 128
+#define MACHINE_ID_LEN 32
+#define UUID_LEN 36
 #define DEFAULT_CMD_SIZE 32
 #define MAX_ID 512
 #define MAX_CGROUP 512
@@ -55,6 +57,8 @@ typedef struct
     char cgroup[MAX_CGROUP];
     char *username;
     char *groupname;
+    char machine_id[MACHINE_ID_LEN + 1];
+    char uuid[UUID_LEN + 1];
 } proc_id_t;
 
 #define TRUE 1

diff --git a/src/utils.c b/src/utils.c
@@ -2,6 +2,7 @@
 #include <errno.h>
 #include <string.h>
 #include <stdlib.h>
+#include <stdio.h>
 #include <time.h>
 
 #include "scopestdlib.h"
@@ -10,6 +11,12 @@
 #include "dbg.h"
 #include "runtimecfg.h"
 
+#define MAC_ADDR_LEN 17
+#define ZERO_MACHINE_ID "00000000000000000000000000000000"
+
+static int createMachineID(char *string);
+static int getMacAddr(char *string);
+
 rtconfig g_cfg = {0};
 
 unsigned int
@@ -203,3 +210,151 @@ sigSafeNanosleep(const struct timespec *req)
 
     return rv;
 }
+
+// Generate a UUID v4 string using a random generator
+void
+setUUID(char *string)
+{
+   if (string == NULL) {
+       scopeLogError("ERROR: setUUIDv4: Null string");
+       return;
+   }
+
+    unsigned char key[16];
+    static bool seeded = FALSE;
+
+    if (!seeded) {
+        scope_srand((unsigned int)scope_time(NULL));
+        seeded = TRUE;
+    }
+
+    for (int i = 0; i < 16; i++) {
+        key[i] = (unsigned char)scope_rand() % 255;
+    }
+
+    key[6] = 0x40 | (key[6] & 0xf); // Set version to 4
+    key[8] = 0x80 | (key[8] & 0x3f); // Set variant to 8
+
+    scope_snprintf(string, UUID_LEN + 1,
+        "%02x%02x%02x%02x-%02x%02x-%02x%02x-%02x%02x-%02x%02x%02x%02x%02x%02x",
+        key[0], key[1], key[2], key[3],
+        key[4], key[5], key[6], key[7], 
+        key[8], key[9], key[10], key[11],
+        key[12], key[13], key[14], key[15]);
+}
+
+// Get the Machine ID, or if not available, create one
+void
+setMachineID(char *string)
+{
+    if (string == NULL) {
+        scopeLogError("ERROR: setMachineID: Null string");
+        return;
+    }
+    scope_strncpy(string, ZERO_MACHINE_ID, MACHINE_ID_LEN + 1);
+
+    char buf[MACHINE_ID_LEN + 1] = {0};
+    FILE *fp;
+
+    // Try to get a machine id from /etc
+    if ((fp = scope_fopen("/etc/machine-id", "r")) != NULL) {
+        if (scope_fgets(buf, sizeof(buf), fp) == NULL) {
+            scopeLogInfo("INFO: setMachineID: Could not read Machine ID from file /etc/machine-id");
+        }
+        scope_fclose(fp);
+    }
+
+    if (scope_strlen(buf) != MACHINE_ID_LEN) {
+        scopeLogInfo("INFO: setMachineID: Machine ID not found or unexpected length. Creating one.");
+        if (createMachineID(buf)) {
+            scopeLogError("ERROR: setMachineID: Error creating Machine ID");
+            return;
+        }
+    }
+
+    scope_strncpy(string, buf, MACHINE_ID_LEN + 1);
+}
+
+// Create a Machine ID from a mac address
+static int
+createMachineID(char *string)
+{
+    if (string == NULL) return 1;
+
+    char mac_addr[MAC_ADDR_LEN];
+    if (getMacAddr(mac_addr)) {
+        scopeLogError("ERROR: createMachineID: getMacAddr");
+        return 1;
+    }
+
+    scope_snprintf(string, MACHINE_ID_LEN + 1, 
+        "%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x%02x",
+        mac_addr[0], mac_addr[1], mac_addr[2], mac_addr[3],
+        mac_addr[4], mac_addr[5], mac_addr[6], mac_addr[7], 
+        mac_addr[8], mac_addr[9], mac_addr[10], mac_addr[11],
+        mac_addr[12], mac_addr[13], mac_addr[14], mac_addr[15]);
+    return 0;
+}
+
+// Get the machine's physical MAC address
+static int
+getMacAddr(char *string)
+{
+    DIR *d;
+    struct dirent *dir;
+    struct stat buf;
+    char mac_buf[MAC_ADDR_LEN];
+    char dir_path[256];
+    char link_path[256];
+    char addr_path[256];
+    bool found = FALSE;
+
+    d = scope_opendir("/sys/class/net/");
+    if (!d) return 1;
+
+    // Check if interface eth exists
+    // Otherwise find an interface that does not contain "virtual" in the soft link
+    while ((dir = scope_readdir(d)) != NULL) {
+        scope_sprintf(dir_path, "/sys/class/net/%s", dir->d_name);
+
+        if (scope_strstr(dir->d_name, "eth") != 0) {
+            found = TRUE;
+            break;
+        }
+
+        if (scope_lstat(dir_path, &buf) != 0) {
+            break;
+        }
+        if (S_ISLNK(buf.st_mode)) {
+            (void)scope_readlink(dir_path, link_path, sizeof(link_path));
+            if (scope_strstr(link_path, "virtual") == NULL) {
+                found = TRUE;
+                break;
+            }
+        }
+    }
+    scope_closedir(d);
+
+    if (!found) {
+        scopeLogError("Error: getMacAddr: No physical interface found");
+        return 1;
+    }
+
+    scope_sprintf(addr_path, "%s/address", dir_path);
+
+    FILE *fp;
+    if ((fp = scope_fopen(addr_path, "r")) == NULL) {
+        scopeLogError("Error: getMacAddr: No address file found");
+        return 1;
+    }
+    if (scope_fgets(mac_buf, sizeof(mac_buf), fp) == NULL) {
+        scopeLogError("Error: getMacAddr: No address found in file");
+        scope_fclose(fp);
+        return 1;
+    }
+    scope_fclose(fp);
+
+    scope_strncpy(string, mac_buf, MAC_ADDR_LEN + 1);
+    return 0;
+}
+
diff --git a/src/utils.h b/src/utils.h
@@ -21,4 +21,7 @@ int endsWith(const char *string, const char *substring);
 
 int sigSafeNanosleep(const struct timespec *req);
 
+void setUUID(char *string);
+void setMachineID(char *string);
+
 #endif // __UTILS_H__
diff --git a/src/wrap.c b/src/wrap.c
@@ -1621,6 +1621,9 @@ init(void)
     setProcId(&g_proc);
     setPidEnv(g_proc.pid);
 
+    setMachineID(g_proc.machine_id);
+    setUUID(g_proc.uuid);
+
     // initEnv() will set this TRUE if it detects `scope_attach_PID.env` in
     // `/dev/shm` with our PID indicating we were injected into a running
     // process.

diff --git a/test/execute.sh b/test/execute.sh
@@ -68,6 +68,7 @@ run_test test/${OS}/cfgutilstest
 run_test test/${OS}/cfgtest
 run_test test/${OS}/transporttest
 run_test test/${OS}/logtest
+run_test test/${OS}/utilstest
 run_test test/${OS}/mtctest
 run_test test/${OS}/evtformattest
 run_test test/${OS}/ctltest