Add capability for check_ceph_health to look a specific checks

This enables the check_ceph_health to select certain monitor health checks
(https://docs.ceph.com/en/latest/rados/operations/health-checks/)
by providing a regexp. This can be used in conjunction with a/the whitelist.

Use-Cases for this might be to have separate checks for certain checks or not
check something specific in one check like flags.

This can also be converted into a blacklist with the right regexp and I provided
an example below how it can be done.

I also add the nagios check command check_ceph_health_filtered with `--check`
and `--whitelist` as arguments as I can imagine it being useful and they play
nicely together. Leaving either argument empty should result in the same result
as omitting the corresponding argument.

Some dummy examples:
$ ./check_ceph_health
WARNING: MON_CLOCK_SKEW( clock skew detected on mon.a )
OBJECT_MISPLACED( 1937172/695961284 objects misplaced (0.278%) )
PG_DEGRADED( Degraded data redundancy: 98/695961284 objects degraded (0.000%), 1 pg degraded )

$ ./check_ceph_health --check 'PG_DEGRADED|OBJECT_MISPLACED'
WARNING: OBJECT_MISPLACED( 1937172/695961284 objects misplaced (0.278%) )
PG_DEGRADED( Degraded data redundancy: 98/695961284 objects degraded (0.000%), 1 pg degraded )

$ ./check_ceph_health --check '^((?!PG_DEGRADED|OBJECT_MISPLACED).)*$'
WARNING: MON_CLOCK_SKEW( clock skew detected on mon.a )

PS: I am not quite happy with the argument name and open to suggestions and
candidates for a short version of the `--check` argument. (`-hc`?)

README.md

@@ -34,6 +34,8 @@ The `check_ceph_health` nagios plugin monitors the ceph cluster, and report its
       -n NAME, --name NAME  ceph client name
       -k KEYRING, --keyring KEYRING
                             ceph client keyring file
+      --check CHECK         regexp of which check(s) to check (luminous+) Can be
+                            inverted, e.g. '^((?!PG_DEGRADED|OBJECT_MISPLACED).)*$'
       -w, --whitelist REGEXP
                             whitelist regexp for ceph health warnings
       -d, --detail          exec 'ceph health detail'
@@ -49,6 +51,19 @@ The `check_ceph_health` nagios plugin monitors the ceph cluster, and report its
 
     nagios$ ./check_ceph_health --id nagios --whitelist 'requests.are.blocked(\s)*32.sec'
 
+    nagios$ ./check_ceph_health --id nagios
+    WARNING: MON_CLOCK_SKEW( clock skew detected on mon.a )
+    OBJECT_MISPLACED( 1937172/695961284 objects misplaced (0.278%) )
+    PG_DEGRADED( Degraded data redundancy: 98/695961284 objects degraded (0.000%), 1 pg degraded )
+
+    nagios$ ./check_ceph_health --id nagios --check 'PG_DEGRADED|OBJECT_MISPLACED'
+    WARNING: OBJECT_MISPLACED( 1937172/695961284 objects misplaced (0.278%) )
+    PG_DEGRADED( Degraded data redundancy: 98/695961284 objects degraded (0.000%), 1 pg degraded )
+
+    nagios$ ./check_ceph_health --id nagios --check '^((?!PG_DEGRADED|OBJECT_MISPLACED).)*$'
+    WARNING: MON_CLOCK_SKEW( clock skew detected on mon.a )
+
+
 ## check_ceph_mon
 
 The `check_ceph_mon` nagios plugin monitors an individual mon daemon, reporting its status.

config/ceph.cfg

@@ -7,6 +7,10 @@ define command{
         command_name    check_ceph_health_wargs
         command_line    /usr/lib/nagios/plugins/check_ceph_health -H '$HOSTADDRESS$'
 }
+define command{
+        command_name    check_ceph_health_filtered
+        command_line    /usr/lib/nagios/plugins/check_ceph_health -H '$HOSTADDRESS$' --check '$ARG1' --whitelist '$ARG2'
+}
 define command{
         command_name    check_ceph_mon
         command_line    /usr/lib/nagios/plugins/check_ceph_mon -I '$ARG1$'

src/check_ceph_health

@@ -23,7 +23,7 @@ import sys
 import re
 import json
 
-__version__ = '1.5.2'
+__version__ = '1.6.0'
 
 # default ceph values
 CEPH_COMMAND = '/usr/bin/ceph'
@@ -45,6 +45,8 @@ def main():
     parser.add_argument('-i','--id', help='ceph client id')
     parser.add_argument('-n','--name', help='ceph client name')
     parser.add_argument('-k','--keyring', help='ceph client keyring file')
+    parser.add_argument('--check', help='regexp of which check(s) to check (luminous+) '
+                        "Can be inverted, e.g. '^((?!PG_DEGRADED|OBJECT_MISPLACED).)*$'")
     parser.add_argument('-w','--whitelist', help='whitelist regexp for ceph health warnings')
     parser.add_argument('-d','--detail', help="exec 'ceph health detail'", action='store_true')
     parser.add_argument('-V','--version', help='show version and exit', action='store_true')
@@ -114,6 +116,10 @@ def main():
         if output.has_key('checks'):
             #luminous
             for check,status in output['checks'].iteritems():
+                # skip check if not selected
+                if args.check and not re.search(args.check, check):
+                    continue
+
                 if status["severity"] == "HEALTH_ERR":
                     extended += msg
                     msg = "CRITCAL: %s( %s )" % (check,status['summary']['message'])