time_t first_t = rrdset_first_entry_t(st);
time_t last_t = rrdset_last_entry_t(st);
- if(unlikely(before - after < st->update_every && after != after - after % st->update_every))
- // when st->update_every is bigger than the frequency we send data to backend
- // skip the iterations that are not aligned to the database
+ if(unlikely(before < first_t || after > last_t))
+ // the chart has not been updated in the wanted timeframe
return NAN;
// align the time-frame
}
static inline int process_graphite_response(BUFFER *b) {
- info("Received %zu bytes from graphite backend. Ignoring them.", buffer_strlen(b));
+ char sample[1024];
+ const char *s = buffer_tostring(b);
+ char *d = sample, *e = &sample[sizeof(sample) - 1];
+
+ for(; *s && d < e ;s++) {
+ char c = *s;
+ if(unlikely(!isprint(c))) c = ' ';
+ *d++ = c;
+ }
+ *d = '\0';
+
+ info("Received %zu bytes from graphite backend. Ignoring them. Sample: '%s'", buffer_strlen(b), sample);
buffer_flush(b);
return 0;
}
static inline int process_opentsdb_response(BUFFER *b) {
- info("Received %zu bytes from opentsdb backend. Ignoring them.", buffer_strlen(b));
+ char sample[1024];
+ const char *s = buffer_tostring(b);
+ char *d = sample, *e = &sample[sizeof(sample) - 1];
+
+ for(; *s && d < e ;s++) {
+ char c = *s;
+ if(unlikely(!isprint(c))) c = ' ';
+ *d++ = c;
+ }
+ *d = '\0';
+
+ info("Received %zu bytes from opentsdb backend. Ignoring them. Sample: '%s'", buffer_strlen(b), sample);
buffer_flush(b);
return 0;
}
void *backends_main(void *ptr) {
- (void)ptr;
+ struct netdata_static_thread *static_thread = (struct netdata_static_thread *)ptr;
BUFFER *b = buffer_create(1), *response = buffer_create(1);
int (*backend_request_formatter)(BUFFER *b, const char *prefix, RRDHOST *host, const char *hostname, RRDSET *st, RRDDIM *rd, time_t after, time_t before, uint32_t options) = NULL;
int default_port = 0;
int sock = -1;
uint32_t options;
- int enabled = config_get_boolean("backend", "enabled", 0);
- const char *source = config_get("backend", "data source", "average");
- const char *type = config_get("backend", "type", "graphite");
+ int enabled = config_get_boolean("backend", "enabled", 0);
+ const char *source = config_get("backend", "data source", "average");
+ const char *type = config_get("backend", "type", "graphite");
const char *destination = config_get("backend", "destination", "localhost");
- const char *prefix = config_get("backend", "prefix", "netdata");
- const char *hostname = config_get("backend", "hostname", localhost.hostname);
- int frequency = (int)config_get_number("backend", "update every", 10);
- int buffer_on_failures = (int)config_get_number("backend", "buffer on failures", 10);
- long timeoutms = config_get_number("backend", "timeout ms", frequency * 2 * 1000);
+ const char *prefix = config_get("backend", "prefix", "netdata");
+ const char *hostname = config_get("backend", "hostname", localhost->hostname);
+ int frequency = (int)config_get_number("backend", "update every", 10);
+ int buffer_on_failures = (int)config_get_number("backend", "buffer on failures", 10);
+ long timeoutms = config_get_number("backend", "timeout ms", frequency * 2 * 1000);
// ------------------------------------------------------------------------
// validate configuration options
// and prepare for sending data to our backend
+
if(!enabled || frequency < 1)
goto cleanup;
goto cleanup;
}
+ if(timeoutms < 1) {
+ error("BACKED invalid timeout %ld ms given. Assuming %d ms.", timeoutms, frequency * 2 * 1000);
+ timeoutms = frequency * 2 * 1000;
+ }
+ timeout.tv_sec = (timeoutms * 1000) / 1000000;
+ timeout.tv_usec = (timeoutms * 1000) % 1000000;
+
+
+ // ------------------------------------------------------------------------
+ // select the backend type
+
if(!strcmp(type, "graphite") || !strcmp(type, "graphite:plaintext")) {
default_port = 2003;
if(options == BACKEND_SOURCE_DATA_AS_COLLECTED)
goto cleanup;
}
- if(timeoutms < 1) {
- error("BACKED invalid timeout %ld ms given. Assuming %d ms.", timeoutms, frequency * 2 * 1000);
- timeoutms = frequency * 2 * 1000;
- }
- timeout.tv_sec = (timeoutms * 1000) / 1000000;
- timeout.tv_usec = (timeoutms * 1000) % 1000000;
// ------------------------------------------------------------------------
- // prepare the charts for monitoring the backend
+ // prepare the charts for monitoring the backend operation
struct rusage thread;
chart_lost_metrics = 0,
chart_sent_metrics = 0,
chart_buffered_bytes = 0,
+ chart_received_bytes = 0,
chart_sent_bytes = 0,
+ chart_receptions = 0,
chart_transmission_successes = 0,
chart_transmission_failures = 0,
chart_data_lost_events = 0,
chart_backend_reconnects = 0,
chart_backend_latency = 0;
- RRDSET *chart_metrics = rrdset_find("netdata.backend_metrics");
- if(!chart_metrics) {
- chart_metrics = rrdset_create("netdata", "backend_metrics", NULL, "backend", NULL, "Netdata Buffered Metrics", "metrics", 130600, frequency, RRDSET_TYPE_LINE);
- rrddim_add(chart_metrics, "buffered", NULL, 1, 1, RRDDIM_ABSOLUTE);
- rrddim_add(chart_metrics, "lost", NULL, 1, 1, RRDDIM_ABSOLUTE);
- rrddim_add(chart_metrics, "sent", NULL, 1, 1, RRDDIM_ABSOLUTE);
- }
-
- RRDSET *chart_bytes = rrdset_find("netdata.backend_bytes");
- if(!chart_bytes) {
- chart_bytes = rrdset_create("netdata", "backend_bytes", NULL, "backend", NULL, "Netdata Backend Data Size", "KB", 130610, frequency, RRDSET_TYPE_AREA);
- rrddim_add(chart_bytes, "buffered", NULL, 1, 1024, RRDDIM_ABSOLUTE);
- rrddim_add(chart_bytes, "lost", NULL, 1, 1024, RRDDIM_ABSOLUTE);
- rrddim_add(chart_bytes, "sent", NULL, 1, 1024, RRDDIM_ABSOLUTE);
- }
-
- RRDSET *chart_ops = rrdset_find("netdata.backend_ops");
- if(!chart_ops) {
- chart_ops = rrdset_create("netdata", "backend_ops", NULL, "backend", NULL, "Netdata Backend Operations", "operations", 130630, frequency, RRDSET_TYPE_LINE);
- rrddim_add(chart_ops, "write", NULL, 1, 1, RRDDIM_ABSOLUTE);
- rrddim_add(chart_ops, "discard", NULL, 1, 1, RRDDIM_ABSOLUTE);
- rrddim_add(chart_ops, "reconnect", NULL, 1, 1, RRDDIM_ABSOLUTE);
- rrddim_add(chart_ops, "failure", NULL, 1, 1, RRDDIM_ABSOLUTE);
- }
-
- RRDSET *chart_latency = rrdset_find("netdata.backend_latency");
- if(!chart_latency) {
- chart_latency = rrdset_create("netdata", "backend_latency", NULL, "backend", NULL, "Netdata Backend Latency", "ms", 130620, frequency, RRDSET_TYPE_AREA);
- rrddim_add(chart_latency, "latency", NULL, 1, 1000, RRDDIM_ABSOLUTE);
- }
+ RRDSET *chart_metrics = rrdset_create_localhost("netdata", "backend_metrics", NULL, "backend", NULL, "Netdata Buffered Metrics", "metrics", 130600, frequency, RRDSET_TYPE_LINE);
+ rrddim_add(chart_metrics, "buffered", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_metrics, "lost", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_metrics, "sent", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+
+ RRDSET *chart_bytes = rrdset_create_localhost("netdata", "backend_bytes", NULL, "backend", NULL, "Netdata Backend Data Size", "KB", 130610, frequency, RRDSET_TYPE_AREA);
+ rrddim_add(chart_bytes, "buffered", NULL, 1, 1024, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_bytes, "lost", NULL, 1, 1024, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_bytes, "sent", NULL, 1, 1024, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_bytes, "received", NULL, 1, 1024, RRD_ALGORITHM_ABSOLUTE);
+
+ RRDSET *chart_ops = rrdset_create_localhost("netdata", "backend_ops", NULL, "backend", NULL, "Netdata Backend Operations", "operations", 130630, frequency, RRDSET_TYPE_LINE);
+ rrddim_add(chart_ops, "write", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_ops, "discard", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_ops, "reconnect", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_ops, "failure", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+ rrddim_add(chart_ops, "read", NULL, 1, 1, RRD_ALGORITHM_ABSOLUTE);
+
+ /*
+ * this is misleading - we can only measure the time we need to send data
+ * this time is not related to the time required for the data to travel to
+ * the backend database and the time that server needed to process them
+ *
+ * issue #1432 and https://www.softlab.ntua.gr/facilities/documentation/unix/unix-socket-faq/unix-socket-faq-2.html
+ *
+ RRDSET *chart_latency = rrdset_create_localhost("netdata", "backend_latency", NULL, "backend", NULL, "Netdata Backend Latency", "ms", 130620, frequency, RRDSET_TYPE_AREA);
+ rrddim_add(chart_latency, "latency", NULL, 1, 1000, RRD_ALGORITHM_ABSOLUTE);
+ */
+
+ RRDSET *chart_rusage = rrdset_create_localhost("netdata", "backend_thread_cpu", NULL, "backend", NULL, "NetData Backend Thread CPU usage", "milliseconds/s", 130630, frequency, RRDSET_TYPE_STACKED);
+ rrddim_add(chart_rusage, "user", NULL, 1, 1000, RRD_ALGORITHM_INCREMENTAL);
+ rrddim_add(chart_rusage, "system", NULL, 1, 1000, RRD_ALGORITHM_INCREMENTAL);
- RRDSET *chart_rusage = rrdset_find("netdata.backend_thread_cpu");
- if(!chart_rusage) {
- chart_rusage = rrdset_create("netdata", "backend_thread_cpu", NULL, "backend", NULL, "NetData Backend Thread CPU usage", "milliseconds/s", 130630, frequency, RRDSET_TYPE_STACKED);
- rrddim_add(chart_rusage, "user", NULL, 1, 1000, RRDDIM_INCREMENTAL);
- rrddim_add(chart_rusage, "system", NULL, 1, 1000, RRDDIM_INCREMENTAL);
- }
// ------------------------------------------------------------------------
// prepare the backend main loop
info("BACKEND configured ('%s' on '%s' sending '%s' data, every %d seconds, as host '%s', with prefix '%s')", type, destination, source, frequency, hostname, prefix);
usec_t step_ut = frequency * USEC_PER_SEC;
- usec_t random_ut = now_realtime_usec() % (step_ut / 2);
- time_t before = (time_t)((now_realtime_usec() - step_ut) / USEC_PER_SEC);
- time_t after = before;
+ time_t after = now_realtime_sec();
int failures = 0;
+ heartbeat_t hb;
+ heartbeat_init(&hb);
for(;;) {
- // ------------------------------------------------------------------------
- // wait for the next iteration point
- usec_t now_ut = now_realtime_usec();
- usec_t next_ut = now_ut - (now_ut % step_ut) + step_ut;
- before = (time_t)(next_ut / USEC_PER_SEC);
+ // ------------------------------------------------------------------------
+ // Wait for the next iteration point.
+ heartbeat_next(&hb, step_ut);
+ time_t before = now_realtime_sec();
- // add a little delay (1/4 of the step) plus some randomness
- next_ut += (step_ut / 4) + random_ut;
-
- while(now_ut < next_ut) {
- sleep_usec(next_ut - now_ut);
- now_ut = now_realtime_usec();
- }
// ------------------------------------------------------------------------
// add to the buffer the data we need to send to the backend
- RRDSET *st;
int pthreadoldcancelstate;
if(unlikely(pthread_setcancelstate(PTHREAD_CANCEL_DISABLE, &pthreadoldcancelstate) != 0))
error("Cannot set pthread cancel state to DISABLE.");
- rrdhost_rdlock(&localhost);
- for(st = localhost.rrdset_root; st ;st = st->next) {
- pthread_rwlock_rdlock(&st->rwlock);
+ RRDHOST *host;
+ for(host = localhost; host ; host = host->next) {
+ // for each host
+
+ rrdhost_rdlock(host);
+
+ RRDSET *st;
+ for(st = host->rrdset_root; st; st = st->next) {
+ // for each chart
- RRDDIM *rd;
- for(rd = st->dimensions; rd ;rd = rd->next) {
- if(rd->last_collected_time.tv_sec >= after)
- chart_buffered_metrics += backend_request_formatter(b, prefix, &localhost, hostname, st, rd, after, before, options);
+ pthread_rwlock_rdlock(&st->rwlock);
+
+ RRDDIM *rd;
+ for(rd = st->dimensions; rd; rd = rd->next) {
+ // for each dimension
+
+ if(rd->last_collected_time.tv_sec >= after)
+ chart_buffered_metrics += backend_request_formatter(b, prefix, host, (host == localhost)?hostname:host->hostname, st, rd, after, before, options);
+ }
+
+ pthread_rwlock_unlock(&st->rwlock);
}
- pthread_rwlock_unlock(&st->rwlock);
+ rrdhost_unlock(host);
}
- rrdhost_unlock(&localhost);
if(unlikely(pthread_setcancelstate(pthreadoldcancelstate, NULL) != 0))
error("Cannot set pthread cancel state to RESTORE (%d).", pthreadoldcancelstate);
chart_buffered_bytes = (collected_number)buffer_strlen(b);
// reset the monitoring chart counters
+ chart_received_bytes =
chart_sent_bytes =
chart_sent_metrics =
chart_lost_metrics =
//fprintf(stderr, "\nBACKEND BEGIN:\n%s\nBACKEND END\n", buffer_tostring(b)); // FIXME
//fprintf(stderr, "after = %lu, before = %lu\n", after, before);
+ // prepare for the next iteration
+ // to add incrementally data to buffer
+ after = before;
+
// ------------------------------------------------------------------------
// if we are connected, receive a response, without blocking
if(likely(sock != -1)) {
- if(likely(fcntl(sock, F_SETFL, O_NONBLOCK) >= 0)) {
- errno = 0;
-
- // loop through to collect all data
- while(sock != -1 && errno != EWOULDBLOCK) {
- buffer_need_bytes(response, 4096);
-
- ssize_t r = recv(sock, &response->buffer[response->len], response->size - response->len
- , O_NONBLOCK);
- if(likely(r > 0)) {
- // we received some data
- response->len += r;
- }
- else if(r == 0) {
- error("Backend '%s' closed the socket", destination);
- close(sock);
- sock = -1;
- }
- else {
- // failed to receive data
- if(errno != EAGAIN && errno != EWOULDBLOCK)
- error("Cannot receive data from backend '%s'.", destination);
- }
+ errno = 0;
+
+ // loop through to collect all data
+ while(sock != -1 && errno != EWOULDBLOCK) {
+ buffer_need_bytes(response, 4096);
+
+ ssize_t r = recv(sock, &response->buffer[response->len], response->size - response->len, MSG_DONTWAIT);
+ if(likely(r > 0)) {
+ // we received some data
+ response->len += r;
+ chart_received_bytes += r;
+ chart_receptions++;
}
-
- if(sock != -1) {
- // unset O_NONBLOCK
- int val = fcntl(sock, F_GETFL, 0);
- if(likely(val >= 0)) {
- int flags = O_NONBLOCK;
- val &= ~flags;
- if(unlikely(fcntl(sock, F_SETFL, val) < 0))
- error("Cannot unset O_NONBLOCK from backend's '%s' socket.", destination);
+ else if(r == 0) {
+ error("Backend '%s' closed the socket", destination);
+ close(sock);
+ sock = -1;
+ }
+ else {
+ // failed to receive data
+ if(errno != EAGAIN && errno != EWOULDBLOCK) {
+ error("Cannot receive data from backend '%s'.", destination);
}
- else
- error("Cannot get active flags from backend's '%s' socket.", destination);
}
-
- // if we received data, process them
- if(buffer_strlen(response))
- backend_response_checker(response);
- }
- else {
- error("Cannot set O_NONBLOCK from backend's '%s' socket. Receiving data from backend skipped.", destination);
}
+
+ // if we received data, process them
+ if(buffer_strlen(response))
+ backend_response_checker(response);
}
// ------------------------------------------------------------------------
// if we are not connected, connect to a backend server
if(unlikely(sock == -1)) {
- usec_t start_ut = now_realtime_usec();
+ usec_t start_ut = now_monotonic_usec();
const char *s = destination;
while(*s) {
const char *e = s;
if(sock != -1) break;
s = e;
}
- chart_backend_latency += now_realtime_usec() - start_ut;
+ chart_backend_latency += now_monotonic_usec() - start_ut;
}
if(unlikely(netdata_exit)) break;
if(likely(sock != -1)) {
size_t len = buffer_strlen(b);
- usec_t start_ut = now_realtime_usec();
+ usec_t start_ut = now_monotonic_usec();
int flags = 0;
#ifdef MSG_NOSIGNAL
flags += MSG_NOSIGNAL;
#endif
ssize_t written = send(sock, buffer_tostring(b), len, flags);
- chart_backend_latency += now_realtime_usec() - start_ut;
+ chart_backend_latency += now_monotonic_usec() - start_ut;
if(written != -1 && (size_t)written == len) {
// we sent the data successfully
chart_transmission_successes++;
close(sock);
sock = -1;
}
-
- // either the buffer is empty
- // or is holding the data we couldn't send
- // so, make sure the next iteration will continue
- // from where we are now
- after = before;
}
else {
error("Failed to update database backend '%s'", destination);
// ------------------------------------------------------------------------
// update the monitoring charts
- if(chart_ops->counter_done) rrdset_next(chart_ops);
+ if(likely(chart_ops->counter_done)) rrdset_next(chart_ops);
+ rrddim_set(chart_ops, "read", chart_receptions);
rrddim_set(chart_ops, "write", chart_transmission_successes);
rrddim_set(chart_ops, "discard", chart_data_lost_events);
rrddim_set(chart_ops, "failure", chart_transmission_failures);
rrddim_set(chart_ops, "reconnect", chart_backend_reconnects);
rrdset_done(chart_ops);
- if(chart_metrics->counter_done) rrdset_next(chart_metrics);
+ if(likely(chart_metrics->counter_done)) rrdset_next(chart_metrics);
rrddim_set(chart_metrics, "buffered", chart_buffered_metrics);
rrddim_set(chart_metrics, "lost", chart_lost_metrics);
rrddim_set(chart_metrics, "sent", chart_sent_metrics);
rrdset_done(chart_metrics);
- if(chart_bytes->counter_done) rrdset_next(chart_bytes);
+ if(likely(chart_bytes->counter_done)) rrdset_next(chart_bytes);
rrddim_set(chart_bytes, "buffered", chart_buffered_bytes);
rrddim_set(chart_bytes, "lost", chart_lost_bytes);
rrddim_set(chart_bytes, "sent", chart_sent_bytes);
+ rrddim_set(chart_bytes, "received", chart_received_bytes);
rrdset_done(chart_bytes);
- if(chart_latency->counter_done) rrdset_next(chart_latency);
+ /*
+ if(likely(chart_latency->counter_done)) rrdset_next(chart_latency);
rrddim_set(chart_latency, "latency", chart_backend_latency);
rrdset_done(chart_latency);
+ */
getrusage(RUSAGE_THREAD, &thread);
- if(chart_rusage->counter_done) rrdset_next(chart_rusage);
+ if(likely(chart_rusage->counter_done)) rrdset_next(chart_rusage);
rrddim_set(chart_rusage, "user", thread.ru_utime.tv_sec * 1000000ULL + thread.ru_utime.tv_usec);
rrddim_set(chart_rusage, "system", thread.ru_stime.tv_sec * 1000000ULL + thread.ru_stime.tv_usec);
rrdset_done(chart_rusage);
info("BACKEND thread exiting");
+ static_thread->enabled = 0;
pthread_exit(NULL);
return NULL;
}
projects / netdata.git / blobdiff