Changeset 3429

Timestamp:

10/07/11 19:32:18 (20 months ago)

Author:

chrfranke

Message:

smartd: Resend warning emails if problem reappears (ticket #167).

Location:

trunk/smartmontools

Files:

• CHANGELOG (modified) (1 diff)
• NEWS (modified) (1 diff)
• smartd.conf.5.in (modified) (5 diffs)
• smartd.cpp (modified) (8 diffs)

trunk/smartmontools/CHANGELOG

@@ -41 +41 @@
 <DEVELOPERS: ADDITIONS TO THE CHANGE LOG GO JUST BELOW HERE, PLEASE>
 
+  [CF] smartd: Resend warning emails if problem reappears (ticket #167).
+
   [CF] smartd: Add separate directives '-l offlinests' and '-l selfteststs'
        to enable tracking of status changes.  Disable '-l offlinests' by

trunk/smartmontools/NEWS

@@ -9 +9 @@
 Summary: smartmontools release 5.42
 -----------------------------------------------------------
+- smartd resends warning emails if problem reappears.
 - smartd directives '-l offlinests' and '-l selfteststs'.
 - Platform-specific man pages.

trunk/smartmontools/smartd.conf.5.in

@@ -597 +597 @@
 
 [ATA only] Failed self-tests outdated by a newer successful extended
-self\-test are ignored.
+self\-test are ignored.  The warning email counter is reset if the
+number of failed self tests dropped to 0.  This typically happens when
+an extended self\-test is run after all bad sectors have been reallocated.
 
 .I offlinests
@@ -847 +849 @@
 type of disk problem detected. Each interval is twice as long as the
 previous interval.
+
+If a disk problem is no longer detected, the internal email counter is
+reset.  If the problem reappears a new warning email is sent immediately.
 
 In addition, one may add zero or more of the following Directives:
@@ -1131 +1136 @@
 See also \'\-v 197,increasing\' below.
 
+The warning email counter is reset if the number of pending sectors
+dropped to 0.  This typically happens when all pending sectors have
+been reallocated or could be read again.
+
 A pending sector is a disk sector (containing 512 bytes of your data)
 which the device would like to mark as ``bad" and reallocate.
@@ -1159 +1168 @@
 See also \'\-v 198,increasing\' below.
 
+The warning email counter is reset if the number of offline uncorrectable
+sectors dropped to 0.  This typically happens when all offline uncorrectable
+sectors have been reallocated or could be read again.
+
 An offline uncorrectable sector is a disk sector which was not
 readable during an off\-line scan or a self\-test.  This is important
@@ -1174 +1187 @@
 will be send if '-m' is specified. If only the limit \fBINFO\fP is
 reached, a message with loglevel \fB\'LOG_INFO\'\fP will be logged.
+
+The warning email counter is reset if the temperature dropped below
+\fBINFO\fP or \fBCRIT\fP-5 if \fBINFO\fP is not specified.
 
 If this directive is used in conjunction with state persistence

trunk/smartmontools/smartd.cpp

@@ -942 +942 @@
     return;
   }
-  
+
   // Return if a single warning mail has been sent.
   if ((cfg.emailfreq==1) && mail->logged)
@@ -1243 +1243 @@
   // increment mail sent counter
   mail->logged++;
+}
+
+static void reset_warning_mail(const dev_config & cfg, dev_state & state, int which, const char *fmt, ...)
+                               __attribute__ ((format (printf, 4, 5)));
+
+static void reset_warning_mail(const dev_config & cfg, dev_state & state, int which, const char *fmt, ...)
+{
+  if (!(0 <= which && which < SMARTD_NMAIL))
+    return;
+
+  // Return if no mail sent yet
+  mailinfo & mi = state.maillog[which];
+  if (!mi.logged)
+    return;
+
+  // Format & print message
+  char msg[256];
+  va_list ap;
+  va_start(ap, fmt);
+  vsnprintf(msg, sizeof(msg), fmt, ap);
+  va_end(ap);
+
+  PrintOut(LOG_INFO, "Device: %s, %s, warning condition reset after %d email%s\n", cfg.name.c_str(),
+           msg, mi.logged, (mi.logged==1 ? "" : "s"));
+
+  // Clear mail counter and timestamps
+  mi = mailinfo();
+  state.must_write = true;
 }
 
@@ -2296 +2324 @@
     // command failed
     MailWarning(cfg, state, 8, "Device: %s, Read SMART Self-Test Log Failed", name);
-  else {      
+  else {
+    reset_warning_mail(cfg, state, 8, "Read SMART Self-Test Log worked again");
+
     // old and new error counts
     int oldc=state.selflogcount;
@@ -2329 +2359 @@
 
     // Print info if error entries have disappeared
-    if (oldc > newc)
+    // or newer successful successful extended self-test exits
+    if (oldc > newc) {
       PrintOut(LOG_INFO, "Device: %s, Self-Test Log error count decreased from %d to %d\n",
                name, oldc, newc);
+      if (newc == 0)
+        reset_warning_mail(cfg, state, 3, "Self-Test Log does no longer report errors");
+    }
 
     // Needed since self-test error count may DECREASE.  Hour might
@@ -2673 +2707 @@
   // No report if no sectors pending.
   uint64_t rawval = ata_get_attr_raw_value(smartval.vendor_attributes[i], cfg.attribute_defs);
-  if (rawval == 0)
+  if (rawval == 0) {
+    reset_warning_mail(cfg, state, mailtype, "No more %s", msg);
     return;
+  }
 
   // If attribute is not reset, report only sector count increases.
@@ -2768 +2804 @@
       cfg.name.c_str(), currtemp, cfg.tempinfo, fmt_temp(state.tempmin, buf), minchg, state.tempmax, maxchg);
   }
+  else if (cfg.tempcrit) {
+    unsigned char limit = (cfg.tempinfo ? cfg.tempinfo : cfg.tempcrit-5);
+    if (currtemp < limit)
+      reset_warning_mail(cfg, state, 12, "Temperature %u Celsius dropped below %u Celsius", currtemp, limit);
+  }
 }
 
@@ -2885 +2926 @@
     MailWarning(cfg, state, 9, "Device: %s, unable to open device", name);
     return 1;
-  } else if (debugmode)
+  }
+  if (debugmode)
     PrintOut(LOG_INFO,"Device: %s, opened ATA device\n", name);
+  reset_warning_mail(cfg, state, 9, "open device worked again");
 
   // user may have requested (with the -n Directive) to leave the disk
@@ -2989 +3032 @@
     }
     else {
+      reset_warning_mail(cfg, state, 6, "read SMART Attribute Data worked again");
+
       // look for current or offline pending sectors
       if (cfg.curr_pending_id)