Checkmk git commits November 2010

checkmk-commits@lists.checkmk.com

2 participants
180 discussions

Check_MK Git: check_mk: Merge branch '1.1.9' of ssh://mathias-kettner.de/ check_mk into 1.1.9

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: d6bc7feb327415ba1176997e9bb7acff8e9324e8 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=d6bc7feb327415… Author: Mathias Kettner <mk(a)mathias-kettner.de> Date: Fri Oct 29 08:35:55 2010 +0200 Merge branch '1.1.9' of ssh://mathias-kettner.de/check_mk into 1.1.9 ---

13 years, 8 months

Check_MK Git: check_mk: Restored check

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: 5d53e496fdf1a75cbcbfe7e7e7bbc271e6998ba1 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=5d53e496fdf1a7… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Fri Oct 29 08:50:51 2010 +0200 Restored check --- checks/cisco_fan | 13 +++++++++---- 1 files changed, 9 insertions(+), 4 deletions(-) diff --git a/checks/cisco_fan b/checks/cisco_fan index b2f522b..04f824c 100644 --- a/checks/cisco_fan +++ b/checks/cisco_fan @@ -24,17 +24,22 @@ # to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, # Boston, MA 02110-1301 USA. +cisco_fan_states = ('', 'normal', 'warning', 'critical', + 'shutdown', 'notPresent', 'notFunctioning') def inventory_cisco_fan(checkname, info): - return [ (line[0], None) for line in info ] + return [ (line[0], None) for line in info if line[1] != '5' ] def check_cisco_fan(item, params, info): for line in info: if line[0] == item: - if line[1] == "1": - return (0, "OK") + state = saveint(line[1]) + if state == 1: + return (0, "OK (State is: %s (%d))" % (cisco_fan_states[state], state)) + elif state == 2: + return (1, "WARNING (state is %s (%d))" % (cisco_fan_states[state], state)) else: - return (2, "CRITICAL (state is %s)" % line[1]) + return (2, "CRITICAL (state is %s (%d))" % (cisco_fan_states[state], state)) return (3, "UNKNOWN - item not found in snmp data") check_info['cisco_fan'] = (check_cisco_fan, "FAN %s", 0, inventory_cisco_fan)

13 years, 8 months

Check_MK Git: check_mk: Merge branch 'master' of git@mathias-kettner.de: check_mk into 1.1.9

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: 1a3884bae98756bb3e9359cb4263cbe7bed011e0 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=1a3884bae98756… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Fri Oct 29 08:49:16 2010 +0200 Merge branch 'master' of git@mathias-kettner.de:check_mk into 1.1.9 Conflicts: checks/cisco_fan ---

13 years, 8 months

Check_MK Git: check_mk: Merge branch 'master' of git@mathias-kettner.de: check_mk into 1.1.9

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: e02fdeccb12b677f66c33a58a53d3f4000983072 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=e02fdeccb12b67… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Wed Oct 27 18:25:10 2010 +0200 Merge branch 'master' of git@mathias-kettner.de:check_mk into 1.1.9 Conflicts: checks/megaraid_pdisks ---

13 years, 8 months

Check_MK Git: check_mk: cleaned up some code in win_dhcp_pools check

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: e48d9c472b48f09a52e2652ea3417baa7d69c057 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=e48d9c472b48f0… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Fri Oct 29 11:20:24 2010 +0200 cleaned up some code in win_dhcp_pools check --- ChangeLog | 1 + LIESMICH.zutun | 19 ------------------- checks/win_dhcp_pools | 36 ++++++++++++------------------------ 3 files changed, 13 insertions(+), 43 deletions(-) diff --git a/ChangeLog b/ChangeLog index bc3c93d..b5551d3 100644 --- a/ChangeLog +++ b/ChangeLog @@ -21,6 +21,7 @@ * blade_health: More detailed output on problems * blade_blades: Added new check for checking the health-, present- and power-state of IBM Bladecenter blades + * win_dhcp_pools: Several cleanups in check Multisite: * The custom open/close states of custom links are now stored for each diff --git a/LIESMICH.zutun b/LIESMICH.zutun index 558032d..429882e 100644 --- a/LIESMICH.zutun +++ b/LIESMICH.zutun @@ -9,25 +9,6 @@ BUGS beheben ab 1.1.9i1 Wenn zwei Checks mit dem gleichen Namen existieren bei einem Host soll mit einem Fehler abgebrochen werden. -win_dhcp_pools sendet: -[['MIBCounts:'], - ['Discovers', '=', '193.'], - ['Offers', '=', '193.'], - ['Delayed', 'Offers', '=', '0.'], - ['Requests', '=', '14540.'], - ['Acks', '=', '11767.'], - ['Naks', '=', '2.'], - ['Declines', '=', '0.'], - ['Releases', '=', '39.'], - ['ServerStartTime', '=', '26', 'June', '2010', '04:47:49'], - ['Scopes', '=', '1.'], - ['Scopes', 'with', 'Delay', 'configured=', '0.'], - ['Subnet', '=', '10.3.108.0.'], - ['No.', 'of', 'Addresses', 'in', 'use', '=', '40.'], - ['No.', 'of', 'free', 'Addresses', '=', '150.'], - ['No.', 'of', 'pending', 'offers', '=', '0.']] -Besser: = als Trenner nehmen, dann ist das parsen einfacher. - ich bin mir nicht sicher ob das nun ein Thruk Problem oder ein Livestatus Problem ist. Vermutlich beides :-) Wenn ich im Thruk im Suchfenster einfach Enter drücke, kommt folgende Query bei raus: GET services diff --git a/checks/win_dhcp_pools b/checks/win_dhcp_pools index bc96e1a..4e5d698 100644 --- a/checks/win_dhcp_pools +++ b/checks/win_dhcp_pools @@ -77,34 +77,26 @@ win_dhcp_pools_stats_translate = { 'Anzahl der anstehenden Angebote': 'No. of pending offers', } +def parse_win_dhcp_pools(info): + return [ ' '.join(line).rstrip('.').split(' = ') for line in info ] + def inventory_win_dhcp_pools(checktype, info): inventory = [] - for line in info: - start = line[0] - if start in win_dhcp_pools_stats_translate: - start = win_dhcp_pools_stats_translate[line[0]] - if start == 'Subnet': - inventory.append((line[2].rstrip('.'), 'win_dhcp_pools_default_levels')) + for line in parse_win_dhcp_pools(info): + if win_dhcp_pools_stats_translate.get(line[0], line[0]) == 'Subnet': + inventory.append((line[1], 'win_dhcp_pools_default_levels')) return inventory def check_win_dhcp_pools(item, params, info): inBlock = False poolStats = [] status = 0 - for line in info: - # Translate if needed - start = line[0] - if start in win_dhcp_pools_stats_translate: - start = win_dhcp_pools_stats_translate[line[0]] - - if start == 'Subnet' and line[-1] == item+'.': + for line in parse_win_dhcp_pools(info): + if win_dhcp_pools_stats_translate.get(line[0], line[0]) == 'Subnet' and line[1] == item: inBlock = True continue if inBlock: - key = ' '.join(line[:-2]) - if key in win_dhcp_pools_stats_translate: - key = win_dhcp_pools_stats_translate[key] - poolStats.append(saveint(line[-1].rstrip('.'))) + poolStats.append(saveint(line[1])) if len(poolStats) == 3: break @@ -140,16 +132,12 @@ def check_win_dhcp_pools_stats(item, params, info): this_time = int(time.time()) timedif = 0 - for line in info: + for line in parse_win_dhcp_pools(info): if len(line) > 0: - if line[0] in win_dhcp_pools_stats_translate: - key = win_dhcp_pools_stats_translate[line[0]] - else: - key = line[0] - + key = win_dhcp_pools_stats_translate.get(line[0], line[0]) if key in [ 'Discovers', 'Offers', 'Requests', 'Acks', 'Nacks', 'Declines', 'Releases', 'Scopes' ]: - value = saveint(line[2].rstrip('.')) + value = saveint(line[1]) try: timedif, per_sec = get_counter("win_dhcp_stats.%s" % key, this_time, value) except MKCounterWrapped:

13 years, 8 months

Check_MK Git: check_mk: blade_*: Added snmp scan functions for better automatic inventory.

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: f4ffd19f864d6bfd95f71624f3cb1c44c7da0bf5 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=f4ffd19f864d6b… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Fri Oct 29 08:40:41 2010 +0200 blade_*: Added snmp scan functions for better automatic inventory. blade_health: More detailed output on problems blade_blades: Added new check for checking the health-, present- and power-state of IBM Bladecenter blades --- ChangeLog | 8 ++++- checkman/blade_bays | 4 +- checks/blade_bays | 3 ++ checks/blade_blades | 77 ++++++++++++++++++++++++++++++++++++++++++++++++ checks/blade_blowers | 3 ++ checks/blade_health | 18 ++++++++--- checks/blade_mediatray | 3 ++ checks/blade_misc | 3 ++ checks/blade_powerfan | 3 ++ checks/blade_powermod | 3 ++ 10 files changed, 116 insertions(+), 9 deletions(-) Diff: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commitdiff;h=f4ffd19f86…

13 years, 8 months

Check_MK Git: check_mk: Fix error detection for tcp+snmp hosts

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: e688626d32701ac5e38c5a43e2d08c8a59094ad5 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=e688626d32701a… Author: Mathias Kettner <mk(a)mathias-kettner.de> Date: Thu Oct 28 11:38:48 2010 +0200 Fix error detection for tcp+snmp hosts --- ChangeLog | 6 ++++ modules/check_mk.py | 11 ++++++- modules/check_mk_base.py | 72 ++++++++++++++++++++++++++++++++++++--------- modules/snmp.py | 4 +- 4 files changed, 74 insertions(+), 19 deletions(-) diff --git a/ChangeLog b/ChangeLog index fc36b90..8f0fd5a 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,4 +1,10 @@ 1.1.9i1 + Core, Setup, etc.: + * Improve error handling: if hosts are monitored with SNMP *and* TCP, + then after an error with one of those two agents checks from the + other haven't been executed. This is fixed now. Inventory check + is still not complete in that error condition. + Checks & Agents: * megaraid_pdisks: Using the real enclosure number as check item now * if/if64: Added expected interface speed to warning output diff --git a/modules/check_mk.py b/modules/check_mk.py index 8a69f09..b1cd68c 100755 --- a/modules/check_mk.py +++ b/modules/check_mk.py @@ -1618,9 +1618,16 @@ def make_inventory(checkname, hostnamelist, check_only=False): try: info = get_realhost_info(hostname, ipaddress, checkname_base, inventory_max_cachefile_age) except MKAgentError, e: - if check_only: + if check_only and str(e): raise - sys.stderr.write("Host '%s': %s\n" % (hostname, str(e))) + elif str(e): + sys.stderr.write("Host '%s': %s\n" % (hostname, str(e))) + continue + except MKSNMPError, e: + if check_only and str(e): + raise + elif str(e): + sys.stderr.write("Host '%s': %s\n" % (hostname, str(e))) continue except Exception, e: if check_only or opt_debug: diff --git a/modules/check_mk_base.py b/modules/check_mk_base.py index 0ba0cfd..e6e719c 100755 --- a/modules/check_mk_base.py +++ b/modules/check_mk_base.py @@ -87,6 +87,8 @@ compiled_regexes = {} # avoid recompiling regexes nagios_command_pipe = None # Filedescriptor to open nagios command pipe. g_single_oid_hostname = None g_single_oid_cache = {} +g_broken_snmp_hosts = set([]) +g_broken_agent_hosts = set([]) # variables set later by getopt @@ -121,6 +123,12 @@ class MKAgentError(Exception): def __str__(self): return self.reason +class MKSNMPError(Exception): + def __init__(self, reason): + self.reason = reason + def __str__(self): + return self.reason + # +----------------------------------------------------------------------+ # | _ _ _ | # | / \ __ _ __ _ _ __ ___ __ _ __ _| |_(_) ___ _ __ | @@ -250,6 +258,7 @@ def get_host_info(hostname, ipaddress, checkname): exception_texts = [] global opt_use_cachefile opt_use_cachefile = True + is_snmp_error = False for node in nodes: # If an error with the agent occurs, we still can (and must) # try the other node. @@ -259,15 +268,23 @@ def get_host_info(hostname, ipaddress, checkname): at_least_one_without_exception = True except MKAgentError, e: exception_texts.append(str(e)) + g_broken_agent_hosts.add(node) + except SNMPErrorError, e: + exception_texts.append(str(e)) + g_broken_snmp_hosts.add(node) + is_snmp_error = true if not at_least_one_without_exception: - raise MKAgentError(", ".join(exception_texts)) + if is_snmp_error: + raise MKSNMPError(", ".join(exception_texts)) + else: + raise MKAgentError(", ".join(exception_texts)) return info else: return get_realhost_info(hostname, ipaddress, checkname, check_max_cachefile_age) # Gets info from a real host (not a cluster). There are three possible # ways: TCP, SNMP and external command. This function raises -# MKAgentError, if there could not retrieved any data. It returns [], +# MKAgentError or MKSNMPError, if there could not retrieved any data. It returns [], # if the agent could be contacted but the data is empty (no items of # this check type). # @@ -290,6 +307,12 @@ def get_realhost_info(hostname, ipaddress, checkname, max_cache_age): content = read_cache_file(cache_relpath, max_cache_age) if content: return eval(content) + # Not cached -> need to get info via SNMP + + # Try to contact host only once + if hostname in g_broken_snmp_hosts: + raise MKSNMPError("") + # New in 1.1.3: oid_info can now be a list: Each element # of that list is interpreted as one real oid_info, fetches # a separate snmp table. The overall result is then the list @@ -320,9 +343,10 @@ def get_realhost_info(hostname, ipaddress, checkname, max_cache_age): return table # No SNMP check. Then we must contact the check_mk_agent. Have we already - # to get data from the agent? If yes we must not do that again! + # to get data from the agent? If yes we must not do that again! Even if + # no cache file is present if g_agent_already_contacted.has_key(hostname): - return [] + raise MKAgentError("") g_agent_already_contacted[hostname] = True store_cached_hostinfo(hostname, []) # leave emtpy info in case of error @@ -390,6 +414,10 @@ def get_agent_info(hostname, ipaddress, max_cache_age): if result: return result + # Try to contact every host only once + if hostname in g_broken_agent_hosts: + raise MKAgentError("") + # If the host ist listed in datasource_programs the data from # that host is retrieved by calling an external program (such # as ssh or rsy) instead of a TCP connect. @@ -625,9 +653,12 @@ def do_check(hostname, ipaddress): try: load_counters(hostname) - agent_version, num_success, num_errors = do_all_checks_on_host(hostname, ipaddress) + agent_version, num_success, num_errors, problems = do_all_checks_on_host(hostname, ipaddress) save_counters(hostname) - if num_errors > 0 and num_success > 0: + if problems: + output = "CRIT - %s" % problems + status = 2 + elif num_errors > 0 and num_success > 0: output = "WARNING - Got only %d out of %d infos" % (num_success, num_success + num_errors) status = 1 elif num_errors > 0: @@ -646,12 +677,6 @@ def do_check(hostname, ipaddress): output = "UNKNOWN - %s" % e status = 3 - except MKAgentError, e: - if opt_debug: - raise - output = "CRIT - %s" % e - status = 2 - if aggregate_check_mk: try: submit_check_mk_aggregation(hostname, status, output) @@ -675,6 +700,7 @@ def do_all_checks_on_host(hostname, ipaddress): num_success = 0 num_errors = 0 check_table = get_sorted_check_table(hostname) + problems = [] for checkname, item, params, description, info in check_table: # In case of a precompiled check table info is the aggrated # service name. In the non-precompiled version there are the dependencies @@ -684,7 +710,22 @@ def do_all_checks_on_host(hostname, ipaddress): aggrname = aggregated_service_name(hostname, description) infotype = checkname.split('.')[0] - info = get_host_info(hostname, ipaddress, infotype) + try: + info = get_host_info(hostname, ipaddress, infotype) + except MKSNMPError, e: + if str(e): + problems.append(str(e)) + num_errors += 1 + g_broken_snmp_hosts.add(hostname) + continue + + except MKAgentError, e: + if str(e): + problems.append(str(e)) + num_errors += 1 + g_broken_agent_hosts.add(hostname) + continue + if info or info == []: num_success += 1 try: @@ -741,10 +782,11 @@ def do_all_checks_on_host(hostname, ipaddress): else: agent_version = "(unknown)" except MKAgentError, e: - raise + g_broken_agent_hosts.add(hostname) + agent_version = "(unknown)" except: agent_version = "(unknown)" - return agent_version, num_success, num_errors + return agent_version, num_success, num_errors, ", ".join(problems) def nagios_pipe_open_timeout(signum, stackframe): raise IOError("Timeout while opening pipe") diff --git a/modules/snmp.py b/modules/snmp.py index a4ac246..ea45083 100644 --- a/modules/snmp.py +++ b/modules/snmp.py @@ -108,7 +108,7 @@ def get_snmp_explicit(hostname, ipaddress, mib, baseoid, suffixes): if exitstatus: if opt_verbose: sys.stderr.write(tty_red + tty_bold + "ERROR: " + tty_normal + "SNMP error\n") - return None + raise MKSNMPError("SNMP Error") return info def snmpwalk_on_suboid(hostname, ip, oid): @@ -153,7 +153,7 @@ def snmpwalk_on_suboid(hostname, ip, oid): if exitstatus: if opt_verbose: sys.stderr.write(tty_red + tty_bold + "ERROR: " + tty_normal + "SNMP error\n") - return [] + raise MKSNMPError("SNMP Error") return rowinfo def get_snmp_table(hostname, ip, oid_info):

13 years, 8 months

Check_MK Git: check_mk: Improved cisco_power check

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: 7b4d7b64214034b33936ddd522db944876d6ef87 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=7b4d7b64214034… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Mon Oct 25 17:14:03 2010 +0200 Improved cisco_power check --- ChangeLog | 5 +++-- LIESMICH.zutun | 2 ++ checks/cisco_power | 13 ++++++++++--- 3 files changed, 15 insertions(+), 5 deletions(-) diff --git a/ChangeLog b/ChangeLog index 5a8ab9a..42177c5 100644 --- a/ChangeLog +++ b/ChangeLog @@ -5,8 +5,9 @@ * wut_webtherm: handle more variants of WuT Webtherms (thanks to Lefty) * blade_bays: Also inventorizes standby blades and has a little more verbose output. - * blade_blowers: Can handle responses without rpm values now. Improved output. - * cisco_fan: Does not inentorize 'notPresent' sensors anymore. Imporved output. + * blade_blowers: Can handle responses without rpm values now. Improved output + * cisco_fan: Does not inentorize 'notPresent' sensors anymore. Improved output + * cisco_power: Not using power source as threshold anymore. Improved output Multisite: * The custom open/close states of custom links are now stored for each diff --git a/LIESMICH.zutun b/LIESMICH.zutun index 3383b0e..27e6ac0 100644 --- a/LIESMICH.zutun +++ b/LIESMICH.zutun @@ -45,6 +45,8 @@ man rechts plötzlich ein anderes System. Lösung ist noch nicht in Sicht. LARS: Bei Opera kann man kein Snapin nach ganz unten ziehen. Der Indikator springt dann immer nach ganz oben. +LARS: Im Opera wird sich die aktuelle Scroll-Position der Sidebar nicht korrekt gemerkt. + SNMP-Checks: Es wurde berichtet, dass - wenn die Community falsch ist - der Check_MK gruen war, aber die Checks unknown. Sollte es dann nicht genau umgekehrt sein? Wird der Exit-code von snmpwalk ausgewertet? diff --git a/checks/cisco_power b/checks/cisco_power index 285c790..25efbc1 100644 --- a/checks/cisco_power +++ b/checks/cisco_power @@ -24,6 +24,9 @@ # to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, # Boston, MA 02110-1301 USA. +cisco_power_states = ('', 'normal', 'warning', 'critical', + 'shutdown', 'notPresent', 'notFunctioning') +cisco_power_source = ( '', 'unknown', 'ac', 'dc', 'externalPowerSupply', 'internalRedundant') def inventory_cisco_power(checkname, info): return [ (line[0], '', '""') for line in info if 'RPS NotExist' not in line[0]] @@ -31,10 +34,14 @@ def inventory_cisco_power(checkname, info): def check_cisco_power(item, params, info): for line in info: if line[0] == item: - if line[1] == "1" and line[2] == "5": - return (0, "OK") + state, source = map(saveint, line[1:3]) + output = 'State: %s, Source: %s' % (cisco_power_states[state], cisco_power_source[source]) + if state == 1: + return (0, "OK - %s" % output) + elif state == 2: + return (1, "WARN - %s" % output) else: - return (2, "CRIT - Invalid state %s/%s" % (line[1], line[2])) + return (2, "CRIT - %s" % output) return (3, "UNKNOWN - item not found in snmp data") check_info['cisco_power'] = (check_cisco_power, "%s", 0, inventory_cisco_power)

13 years, 8 months

Check_MK Git: check_mk: Fixed small coding error in megaraid_pdisks

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: f98b55d493fd26d1477797b65f0744f3ea0fe209 URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=f98b55d493fd26… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Wed Oct 27 11:11:54 2010 +0200 Fixed small coding error in megaraid_pdisks --- checks/megaraid_pdisks | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/checks/megaraid_pdisks b/checks/megaraid_pdisks index 6714f28..9ac8dd0 100644 --- a/checks/megaraid_pdisks +++ b/checks/megaraid_pdisks @@ -65,6 +65,7 @@ def check_megaraid_pdisks(item, target_state, info): # item is "$ENCLOSURE/$SLOT" # param is the target state dev2enc = {} + found = None for line in info: if line[0] == "dev2enc": dev2enc[line[5]] = 'e'+line[2]

13 years, 8 months

Check_MK Git: check_mk: Merge branch 'master' of git@mathias-kettner.de: check_mk into 1.1.9

by git＠mathias-kettner.de

Module: check_mk Branch: master Commit: 5cbebd4fe327fa7b07c7956ae707d7acd8c4333c URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=5cbebd4fe327fa… Author: Lars Michelsen <lm(a)mathias-kettner.de> Date: Tue Oct 26 23:31:00 2010 +0200 Merge branch 'master' of git@mathias-kettner.de:check_mk into 1.1.9 ---

13 years, 8 months

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Checkmk git commits November 2010