System Resilience and Installation Automation Improvements

Makefile

@@ -1,4 +1,4 @@
-.PHONY: all bp_socket daemon clean format check-format
+.PHONY: all bp_socket daemon clean format check-format install
 
 all: bp_socket daemon
 
@@ -15,3 +15,15 @@ clean:
 format:
 	$(MAKE) -C bp_socket format
 	$(MAKE) -C daemon format
+
+install:
+	install -d /usr/local/include
+	install -m 644 include/bp_socket.h /usr/local/include/
+	$(MAKE) -C daemon clean
+	$(MAKE) -C daemon
+	$(MAKE) -C bp_socket clean
+	$(MAKE) -C bp_socket
+	@if lsmod | grep -q "^bp "; then \
+		sudo rmmod bp; \
+	fi
+	sudo insmod bp_socket/bp.ko

Vagrantfile

@@ -24,7 +24,7 @@ Vagrant.configure("2") do |config|
     EOF
     ion.vm.provision "shell", inline: <<-EOF
     export DEBIAN_FRONTEND=noninteractive
-    apt install -y curl git ca-certificates make pkg-config libnl-genl-3-dev libevent-dev build-essential linux-headers-$(uname -r)
+    apt install -y curl git ca-certificates make pkg-config libnl-genl-3-dev libnl-3-dev libevent-dev build-essential clang-format sparse linux-headers-$(uname -r)
 
     cd /opt
     wget -q https://github.com/nasa-jpl/ION-DTN/archive/refs/tags/ion-open-source-4.1.3.tar.gz

bp_client.c

@@ -1,4 +1,4 @@
-#include "include/bp_socket.h"
+#include "bp_socket.h"
 #include <errno.h>
 #include <pthread.h>
 #include <signal.h>
@@ -40,7 +40,8 @@ void *send_thread(void *arg) {
              message_count);
 
     int flags = 0;
-    flags |= MSG_ACK_REQUESTED;
+    // flags |= MSG_ACK_REQUESTED;
+    flags |= MSG_NO_CUSTODY_REQUIRED;
 
     int ret =
         sendto(data->fd, send_buffer, strlen(send_buffer) + 1, flags,

bp_socket/Makefile

@@ -6,7 +6,7 @@ PWD := $(shell pwd)
 
 SRC_FILES := $(wildcard *.c *.h)
 
-.PHONY: all clean sparse
+.PHONY: all clean sparse install
 
 all:
 	$(MAKE) -C /lib/modules/$(shell uname -r)/build M=$(PWD) modules
@@ -18,4 +18,11 @@ sparse:
 	$(MAKE) -C /lib/modules/$(shell uname -r)/build M=$(PWD) modules C=1 CHECK=sparse
 
 format:
-	clang-format -i --style=file $(SRC_FILES)
+	clang-format -i --style=file $(SRC_FILES)
+
+install: all
+	@echo "Installing kernel module..."
+	install -d /lib/modules/$(shell uname -r)/extra
+	install -m 644 bp.ko /lib/modules/$(shell uname -r)/extra/
+	depmod -a
+	@echo "Kernel module installed. Use 'modprobe bp' to load it."

configs/cloud-init/ion.cloud-config.cfg

@@ -10,17 +10,18 @@ packages:
 - make
 - pkg-config 
 - libnl-genl-3-dev 
+- libnl-3-dev
 - libevent-dev
 - build-essential
+- clang-format
+- sparse
 
 ssh_pwauth: false
 users:
 - name: ubuntu
   sudo: ALL=(ALL) NOPASSWD:ALL
   groups: users,admin
   lock_passwd: false
-  ssh_authorized_keys:
-  - "ssh-rsa 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 [email protected]"
   shell: /bin/bash
 
 disable_root: true
@@ -49,8 +50,46 @@ write_files:
     make
     make install
   permissions: '0755'
+- path: /etc/systemd/system/bp-socket-daemon.service
+  content: |
+    [Unit]
+    Description=BP Socket Daemon
+    After=network.target
+
+    [Service]
+    Type=simple
+    User=root
+    WorkingDirectory=/bp-socket
+    Environment=LD_LIBRARY_PATH=/usr/local/lib
+    ExecStart=/bp-socket/daemon/bp_daemon
+    Restart=always
+    RestartSec=5
+
+    [Install]
+    WantedBy=multi-user.target
+  permissions: '0644'
+- path: /etc/systemd/system/ion-dtn.service
+  content: |
+    [Unit]
+    Description=ION DTN Stack
+    After=network.target
+    Wants=network-online.target
+
+    [Service]
+    Type=oneshot
+    User=root
+    WorkingDirectory=/bp-socket/configs
+    Environment=LD_LIBRARY_PATH=/usr/local/lib
+    ExecStart=/usr/local/bin/ionstart -I /bp-socket/configs/host.rc
+    RemainAfterExit=yes
+
+    [Install]
+    WantedBy=multi-user.target
+  permissions: '0644'
 
 runcmd:
 - bash /var/run/scripts/provision.sh
 - apt-get -y install linux-headers-$(uname -r)
-- netplan apply
+- netplan apply
+- cd /bp-socket && make && insmod /bp-socket/bp_socket/bp.ko
+- systemctl daemon-reload

configs/cloud-init/ud3tn.cloud-config.cfg

@@ -20,8 +20,6 @@ users:
   sudo: ALL=(ALL) NOPASSWD:ALL
   groups: users,admin
   lock_passwd: false
-  ssh_authorized_keys:
-  - "ssh-rsa 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 [email protected]"
   shell: /bin/bash
 
 disable_root: true
@@ -51,9 +49,32 @@ write_files:
     make virtualenv
     source .venv/bin/activate
     make update-virtualenv
-  permissions: '0755'
+  permissions: '0755'  
+- path: /etc/systemd/system/ud3tn.service
+  content: |
+    [Unit]
+    Description=uD3TN Bundle Protocol Implementation
+    After=network.target
+
+    [Service]
+    Type=simple
+    User=root
+    WorkingDirectory=/opt/ud3tn
+    ExecStart=/opt/ud3tn/build/posix/ud3tn \
+        --allow-remote-config \
+        --eid ipn:20.0 \
+        --aap2-socket ./ud3tn.aap2.socket.2 \
+        --cla "tcpclv3:*,4556" -L 4
+    Restart=always
+    RestartSec=5
+
+    [Install]
+    WantedBy=multi-user.target
+  permissions: '0644'
 
 runcmd:
 - bash /var/run/scripts/provision.sh
 - chown -R ubuntu:ubuntu /opt/ud3tn
-- netplan apply
+- netplan apply
+- systemctl daemon-reload
+- systemctl enable --now ud3tn.service

daemon/Makefile

@@ -11,7 +11,7 @@ OBJECTS := $(SOURCES:.c=.o)
 
 SRC_FILES := $(wildcard *.c *.h)
 
-.PHONY: all release clean format
+.PHONY: all release clean format install
 
 all: $(EXEC)
 
@@ -28,4 +28,10 @@ clean:
 	rm -f $(EXEC) $(OBJECTS)
 
 format:
-	clang-format -i --style=file $(SRC_FILES)
+	clang-format -i --style=file $(SRC_FILES)
+
+install: $(EXEC)
+	@echo "Installing daemon..."
+	install -d /usr/local/bin
+	install -m 755 $(EXEC) /usr/local/bin/
+	@echo "Daemon installed successfully!"

daemon/bp_genl.c

@@ -72,7 +72,8 @@ int bp_genl_message_handler(struct nl_msg *msg, void *arg) {
     err = nla_parse(attrs, BP_GENL_A_MAX, genlmsg_attrdata(genlhdr, 0), genlmsg_attrlen(genlhdr, 0),
                     NULL);
     if (err < 0) {
-        log_error("Failed to parse Netlink attributes: %s", nl_geterror(err));
+        log_error("bp_genl_handle_msg: failed to parse Netlink attributes for cmd %d: %s",
+                  genlhdr->cmd, nl_geterror(err));
         return NL_SKIP;
     }
 
@@ -86,7 +87,7 @@ int bp_genl_message_handler(struct nl_msg *msg, void *arg) {
     case BP_GENL_CMD_DESTROY_BUNDLE:
         return handle_destroy_bundle(daemon, attrs);
     default:
-        log_error("Unknown Generic Netlink command: %d", genlhdr->cmd);
+        log_error("bp_genl_handle_msg: unknown Generic Netlink command: %d", genlhdr->cmd);
         return NL_SKIP;
     }
 }

daemon/bp_genl_handlers.c

@@ -30,9 +30,10 @@ int handle_open_endpoint(Daemon *daemon, struct nlattr **attrs) {
     ret = ion_open_endpoint(node_id, service_id, daemon->genl_bp_sock, &daemon->netlink_mutex,
                             daemon->genl_bp_family_id);
     if (ret == 0) {
-        log_info("[ipn:%u.%u] OPEN_ENDPOINT: endpoint opened successfully", node_id, service_id);
+        log_info("[ipn:%u.%u] Endpoint opened: spawning receiver and sender threads", node_id,
+                 service_id);
     } else {
-        log_error("[ipn:%u.%u] OPEN_ENDPOINT: failed to open endpoint (error %d)", node_id,
+        log_error("handle_open_endpoint: failed to open endpoint ipn:%u.%u (error %d)", node_id,
                   service_id, ret);
     }
     return ret;
@@ -51,9 +52,9 @@ int handle_close_endpoint(Daemon *daemon, struct nlattr **attrs) {
 
     int ret = ion_close_endpoint(node_id, service_id);
     if (ret == 0) {
-        log_info("[ipn:%u.%u] CLOSE_ENDPOINT: closing endpoint", node_id, service_id);
+        log_info("[ipn:%u.%u] Endpoint closed gracefully", node_id, service_id);
     } else {
-        log_error("[ipn:%u.%u] CLOSE_ENDPOINT: failed to close endpoint (error %d)", node_id,
+        log_error("handle_close_endpoint: failed to close endpoint ipn:%u.%u (error %d)", node_id,
                   service_id, ret);
     }
 
@@ -89,22 +90,19 @@ int handle_send_bundle(Daemon *daemon, struct nlattr **attrs) {
 
     written = snprintf(dest_eid, sizeof(dest_eid), "ipn:%u.%u", dest_node_id, dest_service_id);
     if (written < 0 || written >= (int)sizeof(dest_eid)) {
-        log_error("[ipn:%u.%u] handle_send_bundle: failed to construct EID string", src_node_id,
+        log_error("handle_send_bundle: failed to construct EID string for ipn:%u.%u", src_node_id,
                   src_service_id);
         return -EINVAL;
     }
 
     ret = endpoint_registry_enqueue_send(src_node_id, src_service_id, dest_eid, payload,
                                          payload_size, flags);
     if (ret < 0) {
-        log_error("[ipn:%u.%u] handle_send_bundle: failed to enqueue send (error: %d)", src_node_id,
-                  src_service_id, ret);
+        log_error("handle_send_bundle: failed to enqueue send for ipn:%u.%u (error: %d)",
+                  src_node_id, src_service_id, ret);
         return ret;
     }
 
-    log_info("[ipn:%u.%u] SEND_BUNDLE: bundle queued for sending to EID %s, size %zu (bytes)",
-             src_node_id, src_service_id, dest_eid, payload_size);
-
     return 0;
 }
 
@@ -126,7 +124,7 @@ int handle_destroy_bundle(Daemon *daemon, struct nlattr **attrs) {
         return ret;
     }
 
-    log_info("DESTROY_BUNDLE: bundle consumed by a socket (adu: %llu)", (unsigned long long)adu);
+    log_info("Bundle consumed: successfully destroyed ADU %llu", (unsigned long long)adu);
 
     return 0;
 }

daemon/daemon.c

@@ -44,8 +44,8 @@ int daemon_run(Daemon *self) {
         log_error("Failed to create libevent base");
         return -ENOMEM;
     }
-    log_info("Using libevent version %s with %s behind the scenes", (char *)event_get_version(),
-             (char *)event_base_get_method(self->base));
+    log_debug("Using libevent version %s with %s behind the scenes", (char *)event_get_version(),
+              (char *)event_base_get_method(self->base));
 
     self->event_on_sigint = evsignal_new(self->base, SIGINT, on_sigint, self->base);
     if (!self->event_on_sigint) {
@@ -79,8 +79,6 @@ int daemon_run(Daemon *self) {
         daemon_free(self);
         return -ENOMEM;
     }
-    log_info("Generic Netlink: GENL_BP open socket");
-
     fd = nl_socket_get_fd(self->genl_bp_sock);
     self->event_on_nl_sock = event_new(self->base, fd, EV_READ | EV_PERSIST, on_netlink, self);
     if (!self->event_on_nl_sock) {
@@ -109,9 +107,8 @@ int daemon_run(Daemon *self) {
         return -EAGAIN;
     }
     sdr = bp_get_sdr();
-    log_info("Successfully attached to ION");
 
-    log_info("Daemon started successfully");
+    log_info("Daemon started successfully - attached to ION, Netlink ready");
     event_base_dispatch(self->base);
     log_info("Daemon terminated");
 
@@ -124,13 +121,12 @@ int daemon_run(Daemon *self) {
 void daemon_free(Daemon *self) {
     if (!self) return;
 
-    bp_genl_socket_destroy(self);
-
     if (self->event_on_nl_sock) event_free(self->event_on_nl_sock);
     if (self->event_on_sigpipe) event_free(self->event_on_sigpipe);
     if (self->event_on_sigint) event_free(self->event_on_sigint);
     if (self->base) event_base_free(self->base);
 
+    bp_genl_socket_destroy(self);
     pthread_mutex_destroy(&self->netlink_mutex);
 
 #if LIBEVENT_VERSION_NUMBER >= 0x02010000

daemon/endpoint_registry.c

@@ -150,9 +150,10 @@ int endpoint_registry_enqueue_send(uint32_t node_id, uint32_t service_id, const
 
     pthread_mutex_lock(&ctx->send_queue_mutex);
 
-    if (ctx->send_queue_size >= 1000) { // Limite de queue
+    if (ctx->send_queue_size >= 5000) {
         pthread_mutex_unlock(&ctx->send_queue_mutex);
-        log_warn("endpoint_registry_enqueue_send: queue full for ipn:%u.%u", node_id, service_id);
+        log_warn("endpoint_registry_enqueue_send: queue full for ipn:%u.%u (size: %d)", node_id,
+                 service_id, ctx->send_queue_size);
         free(item->dest_eid);
         free(item->payload);
         free(item);
@@ -172,4 +173,4 @@ int endpoint_registry_enqueue_send(uint32_t node_id, uint32_t service_id, const
     pthread_mutex_unlock(&ctx->send_queue_mutex);
 
     return 0;
-}
+}