kdump support (#729)

In the event of a kernel crash, we need to gather as much information as possible to understand and identify the root cause of the crash. Currently, the kernel does not provide much information, which make kernel crash investigation difficult and time consuming.

Fortunately, there is a way in the kernel to provide more information in the case of a kernel crash. kdump is a feature of the Linux kernel that creates crash dumps in the event of a kernel crash. This PR will add kernel kdump support. Please note that there is another PR in sonic-utilities which is also needed:
sonic-net/sonic-buildimage#3722

An extension to the CLI utilities config and show is provided to configure and manage kdump:

view kdump status (enabled/disabled, active, configuration, stored crash files)
enable / disable kdump functionality
configure kdump (how many kernel crash logs can be saved, memory
allocated for capture kernel)
view kernel crash logs
There is a design document which describes this kdump implementation:
sonic-net/SONiC#510

Author: malletvapid23

config/main.py

@@ -1200,6 +1200,51 @@ def shutdown():
     """Shut down BGP session(s)"""
     pass
 
+@config.group()
+def kdump():
+    """ Configure kdump """
+    if os.geteuid() != 0:
+        exit("Root privileges are required for this operation")
+    pass
+
+@kdump.command()
+def disable():
+    """Disable kdump operation"""
+    config_db = ConfigDBConnector()
+    if config_db is not None:
+        config_db.connect()
+        config_db.mod_entry("KDUMP", "config", {"enabled": "false"})
+        run_command("sonic-kdump-config --disable")
+
+@kdump.command()
+def enable():
+    """Enable kdump operation"""
+    config_db = ConfigDBConnector()
+    if config_db is not None:
+        config_db.connect()
+        config_db.mod_entry("KDUMP", "config", {"enabled": "true"})
+        run_command("sonic-kdump-config --enable")
+
+@kdump.command()
+@click.argument('kdump_memory', metavar='<kdump_memory>', required=True)
+def memory(kdump_memory):
+    """Set memory allocated for kdump capture kernel"""
+    config_db = ConfigDBConnector()
+    if config_db is not None:
+        config_db.connect()
+        config_db.mod_entry("KDUMP", "config", {"memory": kdump_memory})
+        run_command("sonic-kdump-config --memory %s" % kdump_memory)
+
+@kdump.command()
+@click.argument('kdump_num_dumps', metavar='<kdump_num_dumps>', required=True, type=int)
+def num_dumps(kdump_num_dumps):
+    """Set max number of dump files for kdump"""
+    config_db = ConfigDBConnector()
+    if config_db is not None:
+        config_db.connect()
+        config_db.mod_entry("KDUMP", "config", {"num_dumps": kdump_num_dumps})
+        run_command("sonic-kdump-config --num_dumps %d" % kdump_num_dumps)
+
 # 'all' subcommand
 @shutdown.command()
 @click.option('-v', '--verbose', is_flag=True, help="Enable verbose output")

scripts/generate_dump

@@ -436,6 +436,18 @@ main() {
         fi
     done
 
+    # archive kernel dump files
+    for file in $(find_files "/var/crash/"); do
+        # don't gzip already-gzipped dmesg files :)
+        if [ ! ${file} = "/var/crash/kexec_cmd" -a ! ${file} = "/var/crash/export" ]; then
+            if [[ ${file} == *"kdump."* ]]; then
+                save_file $file kdump false
+            else
+                save_file $file kdump true
+            fi
+        fi
+    done
+
     # clean up working tar dir before compressing
     $RM $V -rf $TARDIR
 

scripts/reboot

@@ -1,5 +1,12 @@
 #!/bin/bash
 
+# Reboot immediately if we run the kdump capture kernel
+VMCORE_FILE=/proc/vmcore
+if [ -e $VMCORE_FILE -a -s $VMCORE_FILE ]; then
+        debug "We have a /proc/vmcore, then we just kdump'ed"
+        /sbin/reboot
+fi
+
 REBOOT_USER=$(logname)
 REBOOT_TIME=$(date)
 PLATFORM=$(sonic-cfggen -H -v DEVICE_METADATA.localhost.platform)