#!/usr/bin/env python # # Copyright (c) 2013, Antonio Verni, me.verni@gmail.com # # Permission is hereby granted, free of charge, to any person obtaining a # copy of this software and associated documentation files (the "Software"), # to deal in the Software without restriction, including without limitation # the rights to use, copy, modify, merge, publish, distribute, sublicense, # and/or sell copies of the Software, and to permit persons to whom the # Software is furnished to do so, subject to the following conditions: # # The above copyright notice and this permission notice shall be included # in all copies or substantial portions of the Software. # # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS # OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL # THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING # FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER # DEALINGS IN THE SOFTWARE. # # Solr 4.* munin graph plugin # Plugins configuration parameters: # # [solr_*] # env.host_port # env.url # env.qpshandler_ # # ex: # env.host_port solrhost:8080 # env.url /solr # env.qpshandler_select /select # # Install plugins: # ln -s /usr/share/munin/plugins/solr_.py /etc/munin/plugins/solr_numdocs_core_1 # ln -s /usr/share/munin/plugins/solr_.py /etc/munin/plugins/solr_requesttimes_select # ln -s /usr/share/munin/plugins/solr_.py /etc/munin/plugins/solr_core_1_select # ln -s /usr/share/munin/plugins/solr_.py /etc/munin/plugins/solr_indexsize # ln -s /usr/share/munin/plugins/solr_.py /etc/munin/plugins/solr_memory # import sys import os import httplib import json def parse_params(): plugname = os.path.basename(sys.argv[0]).split('_', 2)[1:] params = { 'type': plugname[0], 'op': 'config' if sys.argv[-1] == 'config' else 'fetch', 'core': plugname[1] if len(plugname) > 1 else '', 'params': {} } if plugname[0] in[ 'qps', 'requesttimes']: data = params['core'].rsplit('_', 1) handler = data.pop() params['params'] = { 'handler': os.environ.get('qpshandler_%s' % handler, 'standard') } if not data: params['core'] = '' else: params['core'] = data[0] elif plugname[0] == 'indexsize': params['params']['core'] = params['core'] return params ############################################################################# # Datasources class CheckException(Exception): pass class JSONReader: @classmethod def readValue(cls, struct, path, convert = None): if not path[0] in struct: return -1 obj = struct[path[0]] if not obj: return -1 for k in path[1:]: obj = obj[k] if convert: return convert(obj) return obj class SolrCoresAdmin: def __init__(self, host, solrurl): self.host = host self.solrurl = solrurl self.data = None def fetchcores(self): uri = os.path.join(self.solrurl, "admin/cores?action=STATUS&wt=json") conn = httplib.HTTPConnection(self.host) conn.request("GET", uri) res = conn.getresponse() data = res.read() if res.status != 200: raise CheckException("Cores status fetch failed: %s\n%s" %( str(res.status), res.read())) self.data = json.loads(data) def getCores(self): if not self.data: self.fetchcores() cores = JSONReader.readValue(self.data, ['status']) return cores.keys() def indexsize(self, core = None): if not self.data: self.fetchcores() if core: return { core: JSONReader.readValue(self.data, ['status', core, 'index', 'sizeInBytes']) } else: ret = {} for core in self.getCores(): ret[core] = JSONReader.readValue(self.data, ['status', core, 'index', 'sizeInBytes']) return ret class SolrCoreMBean: def __init__(self, host, solrurl, core): self.host = host self.data = None self.core = core self.solrurl = solrurl def _fetch(self): uri = os.path.join(self.solrurl, "%s/admin/mbeans?stats=true&wt=json" % self.core) conn = httplib.HTTPConnection(self.host) conn.request("GET", uri) res = conn.getresponse() data = res.read() if res.status != 200: raise CheckException("MBean fetch failed: %s\n%s" %( str(res.status), res.read())) raw_data = json.loads(data) data = {} self.data = { 'solr-mbeans': data } key = None for pos, el in enumerate(raw_data['solr-mbeans']): if pos % 2 == 1: data[key] = el else: key = el self._fetchSystem() def _fetchSystem(self): uri = os.path.join(self.solrurl, "%s/admin/system?stats=true&wt=json" % self.core) conn = httplib.HTTPConnection(self.host) conn.request("GET", uri) res = conn.getresponse() data = res.read() if res.status != 200: raise CheckException("System fetch failed: %s\n%s" %( str(res.status), res.read())) self.data['system'] = json.loads(data) def _readInt(self, path): return self._read(path, int) def _readFloat(self, path): return self._read(path, float) def _read(self, path, convert = None): if self.data is None: self._fetch() return JSONReader.readValue(self.data, path, convert) def _readCache(self, cache): result = {} for key, ftype in [('lookups', int), ('hits', int), ('inserts', int), ('evictions', int), ('hitratio', float)]: path = ['solr-mbeans', 'CACHE', cache, 'stats', 'cumulative_%s' % key] result[key] = self._read(path, ftype) result['size'] = self._readInt(['solr-mbeans', 'CACHE', cache, 'stats', 'size']) return result def getCore(self): return self.core def requestcount(self, handler): path = ['solr-mbeans', 'QUERYHANDLER', handler, 'stats', 'requests'] return self._readInt(path) def qps(self, handler): path = ['solr-mbeans', 'QUERYHANDLER', handler, 'stats', 'avgRequestsPerSecond'] return self._readFloat(path) def requesttimes(self, handler): times = {} path = ['solr-mbeans', 'QUERYHANDLER', handler, 'stats'] for perc in ['avgTimePerRequest', '75thPcRequestTime', '99thPcRequestTime']: times[perc] = self._read(path + [perc], float) return times def numdocs(self): path = ['solr-mbeans', 'CORE', 'searcher', 'stats', 'numDocs'] return self._readInt(path) def documentcache(self): return self._readCache('documentCache') def filtercache(self): return self._readCache('filterCache') def fieldvaluecache(self): return self._readCache('fieldValueCache') def queryresultcache(self): return self._readCache('queryResultCache') def memory(self): data = self._read(['system', 'jvm', 'memory', 'raw']) del data['used%'] for k in data.keys(): data[k] = int(data[k]) return data ############################################################################# # Graph Templates CACHE_GRAPH_TPL = """multigraph solr_{core}_{cacheType}_hit_rates graph_category solr graph_title Solr {core} {cacheName} Hit rates graph_order lookups hits inserts graph_scale no graph_vlabel Hit Rate graph_args -u 100 --rigid lookups.label Cache lookups lookups.graph no lookups.min 0 lookups.type DERIVE inserts.label Cache misses inserts.min 0 inserts.draw STACK inserts.cdef inserts,lookups,/,100,* inserts.type DERIVE hits.label Cache hits hits.min 0 hits.draw AREA hits.cdef hits,lookups,/,100,* hits.type DERIVE multigraph solr_{core}_{cacheType}_size graph_title Solr {core} {cacheName} Size graph_args -l 0 graph_category solr graph_vlabel Size size.label Size size.draw LINE2 evictions.label Evictions evictions.draw LINE2 """ QPSMAIN_GRAPH_TPL = """graph_title Solr {core} {handler} Request per second graph_args --base 1000 -r --lower-limit 0 graph_scale no graph_vlabel request / second graph_category solr graph_period second graph_order {gorder} {cores_qps_graphs}""" QPSCORE_GRAPH_TPL = """qps_{core}.label {core} Request per second qps_{core}.draw {gtype} qps_{core}.type DERIVE qps_{core}.min 0 qps_{core}.graph yes""" REQUESTTIMES_GRAPH_TPL = """multigraph {core}_requesttimes graph_title Solr {core} {handler} Time per request graph_args -l 0 graph_vlabel millis graph_category solr savgtimeperrequest_{core}.label {core} Avg time per request savgtimeperrequest_{core}.type GAUGE savgtimeperrequest_{core}.graph yes s75thpcrequesttime_{core}.label {core} 75th perc s75thpcrequesttime_{core}.type GAUGE s75thpcrequesttime_{core}.graph yes s99thpcrequesttime_{core}.label {core} 99th perc s99thpcrequesttime_{core}.type GAUGE s99thpcrequesttime_{core}.graph yes """ NUMDOCS_GRAPH_TPL = """graph_title Solr Docs %s graph_vlabel docs docs.label Docs graph_category solr""" INDEXSIZE_GRAPH_TPL = """graph_args --base 1024 -l 0 graph_vlabel Bytes graph_title Index Size graph_category solr graph_info Solr Index Size. graph_order {cores} {cores_config} xmx.label Xmx xmx.colour ff0000 """ INDEXSIZECORE_GRAPH_TPL = """{core}.label {core} {core}.draw STACK""" MEMORYUSAGE_GRAPH_TPL = """graph_args --base 1024 -l 0 --upper-limit {availableram} graph_vlabel Bytes graph_title Solr memory usage graph_category solr graph_info Solr Memory Usage. used.label Used max.label Max max.colour ff0000 """ ############################################################################# # Graph managment class SolrMuninGraph: def __init__(self, hostport, solrurl, params): self.solrcoresadmin = SolrCoresAdmin(hostport, solrurl) self.hostport = hostport self.solrurl = solrurl self.params = params def _getMBean(self, core): return SolrCoreMBean(self.hostport, self.solrurl, core) def _cacheConfig(self, cacheType, cacheName): return CACHE_GRAPH_TPL.format(core=self.params['core'], cacheType=cacheType, cacheName=cacheName) def _format4Value(self, value): if isinstance(value, basestring): return "%s" if isinstance(value, int): return "%d" if isinstance(value, float): return "%.6f" return "%s" def _cacheFetch(self, cacheType, fields = None): fields = fields or ['size', 'lookups', 'hits', 'inserts', 'evictions'] hits_fields = ['lookups', 'hits', 'inserts'] size_fields = ['size', 'evictions'] results = [] solrmbean = self._getMBean(self.params['core']) data = getattr(solrmbean, cacheType)() results.append('multigraph solr_{core}_{cacheType}_hit_rates'.format(core=self.params['core'], cacheType=cacheType)) for label in hits_fields: vformat = self._format4Value(data[label]) results.append(("%s.value " + vformat) % (label, data[label])) results.append('multigraph solr_{core}_{cacheType}_size'.format(core=self.params['core'], cacheType=cacheType)) for label in size_fields: results.append("%s.value %d" % (label, data[label])) return "\n".join(results) def config(self, mtype): if not mtype or not hasattr(self, '%sConfig' % mtype): raise CheckException("Unknown check %s" % mtype) return getattr(self, '%sConfig' % mtype)() def fetch(self, mtype): if not hasattr(self, params['type']): return None return getattr(self, params['type'])() def _getCores(self): if self.params['core']: cores = [self.params['core']] else: cores = sorted(self.solrcoresadmin.getCores()) return cores def qpsConfig(self): cores = self._getCores() graph = [QPSCORE_GRAPH_TPL.format(core=c, gtype='LINESTACK1') for pos,c in enumerate(cores) ] return QPSMAIN_GRAPH_TPL.format( cores_qps_graphs='\n'.join(graph), handler=self.params['params']['handler'], core=self.params['core'], cores_qps_cdefs='%s,%s' % (','.join(map(lambda x: 'qps_%s' % x, cores)),','.join(['+']*(len(cores)-1))), gorder=','.join(cores) ) def qps(self): results = [] cores = self._getCores() for c in cores: mbean = self._getMBean(c) results.append('qps_%s.value %d' % (c, mbean.requestcount(self.params['params']['handler']))) return '\n'.join(results) def requesttimesConfig(self): cores = self._getCores() graphs = [REQUESTTIMES_GRAPH_TPL.format(core=c, handler=self.params['params']['handler']) for c in cores ] return '\n'.join(graphs) def requesttimes(self): cores = self._getCores() results = [] for c in cores: mbean = self._getMBean(c) results.append('multigraph {core}_requesttimes'.format(core=c)) for k, time in mbean.requesttimes(self.params['params']['handler']).items(): results.append('s%s_%s.value %.5f' % (k.lower(), c, time)) return '\n'.join(results) def numdocsConfig(self): return NUMDOCS_GRAPH_TPL % self.params['core'] def numdocs(self): mbean = self._getMBean(self.params['core']) return 'docs.value %d' % mbean.numdocs(**self.params['params']) def indexsizeConfig(self): cores = self._getCores() graph = [ INDEXSIZECORE_GRAPH_TPL.format(core=c) for c in cores] return INDEXSIZE_GRAPH_TPL.format(cores=" ".join(cores), cores_config="\n".join(graph)) def indexsize(self): results = [] for c, size in self.solrcoresadmin.indexsize(**self.params['params']).items(): results.append("%s.value %d" % (c, size)) cores = self._getCores() mbean = self._getMBean(cores[0]) memory = mbean.memory() results.append('xmx.value %d' % memory['max']) return "\n".join(results) def memoryConfig(self): cores = self._getCores() mbean = self._getMBean(cores[0]) memory = mbean.memory() return MEMORYUSAGE_GRAPH_TPL.format(availableram=memory['max'] * 1.05) def memory(self): results = [] cores = self._getCores() mbean = self._getMBean(cores[0]) memory = mbean.memory() return '\n'.join(['used.value %d' % memory['used'], 'max.value %d' % memory['max']]) def documentcacheConfig(self): return self._cacheConfig('documentcache', 'Document Cache') def documentcache(self): return self._cacheFetch('documentcache') def filtercacheConfig(self): return self._cacheConfig('filtercache', 'Filter Cache') def filtercache(self): return self._cacheFetch('filtercache') def fieldvaluecacheConfig(self): return self._cacheConfig('fieldvaluecache', 'Field Value Cache') def fieldvaluecache(self): return self._cacheFetch('fieldvaluecache') def queryresultcacheConfig(self): return self._cacheConfig('queryresultcache', 'Query Cache') def queryresultcache(self): return self._cacheFetch('queryresultcache') if __name__ == '__main__': params = parse_params() SOLR_HOST_PORT = os.environ.get('host_port', 'localhost:8080').replace('http://', '') SOLR_URL = os.environ.get('url', '/solr') if SOLR_URL[0] != '/': SOLR_URL = '/' + SOLR_URL mb = SolrMuninGraph(SOLR_HOST_PORT, SOLR_URL, params) if hasattr(mb, params['op']): print getattr(mb, params['op'])(params['type'])