Context Navigation

NSClient++ Help (#1) - Missing NSCA results [SOLVED] (#618) - Message List

Missing NSCA results [SOLVED]

sorry this is so long...errored on the side of too much info

We are missing NSCA results. During testing of some test templates, approx 74 checks in 4 different ini files, we noticed we weren't see all the results on the servers nsca log (syslog). We tracket it back and determine it is a client issue. Here are some facts.

tcp captures confirm the client is not sending all results, in this case all 74. Captures done on both client and server.

Confirmed by adding 300 simple checks to one ini file. The client is only sending ~32 results per interval, interval set to 60 while testing

/test not showing any errors. It indicates all tests are ran and injected

When testing 74 checks and the 300 test checks they client finished <5seconds

I have disabled encryption, still happening

Testing shows that the client typically isn't sending more than ~65 per interval. The results that are sent do seem to differ. So for the 1-300 test (NG-TC-WIN-BASE_1..300) I broke these down into the 1 min buckets since our interval is 1 min. As you can see the first bunch of results are the same everytime then it seems to send various blocks of results.

Please help because this is a show stopper for us!

14:28:50-'NG-TC-W2K3-BASE_1',
14:28:50-'NG-TC-W2K3-BASE_10',
14:28:50-'NG-TC-W2K3-BASE_100',
14:28:50-'NG-TC-W2K3-BASE_101',
14:28:50-'NG-TC-W2K3-BASE_102',
14:28:50-'NG-TC-W2K3-BASE_103',
14:28:50-'NG-TC-W2K3-BASE_104',
14:28:50-'NG-TC-W2K3-BASE_105',
14:28:50-'NG-TC-W2K3-BASE_106',
14:28:50-'NG-TC-W2K3-BASE_107',
14:28:50-'NG-TC-W2K3-BASE_108',
14:28:50-'NG-TC-W2K3-BASE_109',
14:28:50-'NG-TC-W2K3-BASE_112',
14:28:50-'NG-TC-W2K3-BASE_113',
14:28:50-'NG-TC-W2K3-BASE_114',
14:28:50-'NG-TC-W2K3-BASE_115',
14:28:50-'NG-TC-W2K3-BASE_117',
14:28:50-'NG-TC-W2K3-BASE_118',
14:28:50-'NG-TC-W2K3-BASE_119',
14:28:50-'NG-TC-W2K3-BASE_12',
14:28:50-'NG-TC-W2K3-BASE_120',
14:28:50-'NG-TC-W2K3-BASE_121',
14:28:50-'NG-TC-W2K3-BASE_122',
14:28:50-'NG-TC-W2K3-BASE_123',
14:28:50-'NG-TC-W2K3-BASE_124',
14:28:50-'NG-TC-W2K3-BASE_125',
14:28:50-'NG-TC-W2K3-BASE_126',
14:28:50-'NG-TC-W2K3-BASE_127',
14:28:50-'NG-TC-W2K3-BASE_128',
14:28:50-'NG-TC-W2K3-BASE_129',
14:28:50-'NG-TC-W2K3-BASE_13',
14:28:50-'NG-TC-W2K3-BASE_130',
14:28:50-'NG-TC-W2K3-BASE_131',
14:28:50-'NG-TC-W2K3-BASE_132',
14:28:50-'NG-TC-W2K3-BASE_133',
14:28:50-'NG-TC-W2K3-BASE_134',
14:28:50-'NG-TC-W2K3-BASE_135',
14:28:50-'NG-TC-W2K3-BASE_136',
14:28:50-'NG-TC-W2K3-BASE_137',
14:28:50-'NG-TC-W2K3-BASE_138',
14:28:50-'NG-TC-W2K3-BASE_139',
14:28:50-'NG-TC-W2K3-BASE_14',
14:28:50-'NG-TC-W2K3-BASE_140',
14:28:50-'NG-TC-W2K3-BASE_141',
14:28:50-'NG-TC-W2K3-BASE_142',
total of 46
----
14:29:51-'NG-TC-W2K3-BASE_1',
14:29:51-'NG-TC-W2K3-BASE_10',
14:29:51-'NG-TC-W2K3-BASE_100',
14:29:51-'NG-TC-W2K3-BASE_101',
14:29:51-'NG-TC-W2K3-BASE_102',
14:29:51-'NG-TC-W2K3-BASE_103',
14:29:51-'NG-TC-W2K3-BASE_104',
14:29:51-'NG-TC-W2K3-BASE_105',
14:29:51-'NG-TC-W2K3-BASE_106',
14:29:51-'NG-TC-W2K3-BASE_107',
14:29:51-'NG-TC-W2K3-BASE_108',
14:29:51-'NG-TC-W2K3-BASE_109',
14:29:51-'NG-TC-W2K3-BASE_113',
14:29:51-'NG-TC-W2K3-BASE_114',
14:29:51-'NG-TC-W2K3-BASE_115',
14:29:51-'NG-TC-W2K3-BASE_116',
14:29:51-'NG-TC-W2K3-BASE_126',
14:29:51-'NG-TC-W2K3-BASE_127',
14:29:51-'NG-TC-W2K3-BASE_128',
14:29:51-'NG-TC-W2K3-BASE_129',
14:29:51-'NG-TC-W2K3-BASE_13',
14:29:51-'NG-TC-W2K3-BASE_130',
14:29:51-'NG-TC-W2K3-BASE_131',
14:29:51-'NG-TC-W2K3-BASE_132',
14:29:51-'NG-TC-W2K3-BASE_133',
14:29:51-'NG-TC-W2K3-BASE_134',
14:29:51-'NG-TC-W2K3-BASE_135',
14:29:51-'NG-TC-W2K3-BASE_136',
14:29:51-'NG-TC-W2K3-BASE_137',
14:29:51-'NG-TC-W2K3-BASE_138',
14:29:51-'NG-TC-W2K3-BASE_139',
14:29:51-'NG-TC-W2K3-BASE_14',
14:29:51-'NG-TC-W2K3-BASE_140',
14:29:51-'NG-TC-W2K3-BASE_141',
14:29:51-'NG-TC-W2K3-BASE_142',
14:29:51-'NG-TC-W2K3-BASE_143',
14:29:51-'NG-TC-W2K3-BASE_144',
14:29:51-'NG-TC-W2K3-BASE_147',
14:29:51-'NG-TC-W2K3-BASE_148',
14:29:51-'NG-TC-W2K3-BASE_158',
14:29:51-'NG-TC-W2K3-BASE_159',
14:29:51-'NG-TC-W2K3-BASE_16',
14:29:51-'NG-TC-W2K3-BASE_160',
14:29:51-'NG-TC-W2K3-BASE_161',
14:29:51-'NG-TC-W2K3-BASE_162',
14:29:51-'NG-TC-W2K3-BASE_39',
14:29:51-'NG-TC-W2K3-BASE_4',
14:29:51-'NG-TC-W2K3-BASE_40',
14:29:51-'NG-TC-W2K3-BASE_41',
14:29:51-'NG-TC-W2K3-BASE_42',
14:29:51-'NG-TC-W2K3-BASE_43',
14:29:51-'NG-TC-W2K3-BASE_44',
14:29:51-'NG-TC-W2K3-BASE_45',
14:29:51-'NG-TC-W2K3-BASE_46',
14:29:51-'NG-TC-W2K3-BASE_47',
14:29:51-'NG-TC-W2K3-BASE_48',
14:29:51-'NG-TC-W2K3-BASE_49',
total 58
----
14:30:50-'NG-TC-W2K3-BASE_1',
14:30:50-'NG-TC-W2K3-BASE_10',
14:30:50-'NG-TC-W2K3-BASE_100',
14:30:50-'NG-TC-W2K3-BASE_101',
14:30:50-'NG-TC-W2K3-BASE_102',
14:30:50-'NG-TC-W2K3-BASE_103',
14:30:50-'NG-TC-W2K3-BASE_104',
14:30:50-'NG-TC-W2K3-BASE_105',
14:30:50-'NG-TC-W2K3-BASE_106',
14:30:50-'NG-TC-W2K3-BASE_107',
14:30:50-'NG-TC-W2K3-BASE_108',
14:30:50-'NG-TC-W2K3-BASE_109',
14:30:50-'NG-TC-W2K3-BASE_111',
14:30:50-'NG-TC-W2K3-BASE_113',
14:30:50-'NG-TC-W2K3-BASE_114',
14:30:50-'NG-TC-W2K3-BASE_115',
14:30:50-'NG-TC-W2K3-BASE_116',
14:30:50-'NG-TC-W2K3-BASE_117',
14:30:50-'NG-TC-W2K3-BASE_119',
14:30:50-'NG-TC-W2K3-BASE_12',
14:30:50-'NG-TC-W2K3-BASE_120',
14:30:50-'NG-TC-W2K3-BASE_121',
14:30:50-'NG-TC-W2K3-BASE_122',
14:30:50-'NG-TC-W2K3-BASE_123',
14:30:50-'NG-TC-W2K3-BASE_124',
14:30:50-'NG-TC-W2K3-BASE_125',
14:30:50-'NG-TC-W2K3-BASE_126',
14:30:50-'NG-TC-W2K3-BASE_127',
14:30:50-'NG-TC-W2K3-BASE_128',
14:30:50-'NG-TC-W2K3-BASE_129',
14:30:50-'NG-TC-W2K3-BASE_13',
14:30:50-'NG-TC-W2K3-BASE_130',
14:30:50-'NG-TC-W2K3-BASE_131',
14:30:50-'NG-TC-W2K3-BASE_132',
14:30:50-'NG-TC-W2K3-BASE_133',
14:30:50-'NG-TC-W2K3-BASE_134',
14:30:50-'NG-TC-W2K3-BASE_135',
14:30:50-'NG-TC-W2K3-BASE_136',
14:30:50-'NG-TC-W2K3-BASE_137',
14:30:50-'NG-TC-W2K3-BASE_138',
14:30:50-'NG-TC-W2K3-BASE_139',
14:30:50-'NG-TC-W2K3-BASE_14',
14:30:50-'NG-TC-W2K3-BASE_140',
14:30:50-'NG-TC-W2K3-BASE_254',
14:30:50-'NG-TC-W2K3-BASE_255',
14:30:50-'NG-TC-W2K3-BASE_256',
14:30:50-'NG-TC-W2K3-BASE_257',
14:30:50-'NG-TC-W2K3-BASE_258',
14:30:50-'NG-TC-W2K3-BASE_259',
14:30:50-'NG-TC-W2K3-BASE_26',
14:30:50-'NG-TC-W2K3-BASE_260',
14:30:50-'NG-TC-W2K3-BASE_261',
14:30:50-'NG-TC-W2K3-BASE_262',
14:30:50-'NG-TC-W2K3-BASE_263',
14:30:50-'NG-TC-W2K3-BASE_264',
14:30:50-'NG-TC-W2K3-BASE_53',
14:30:50-'NG-TC-W2K3-BASE_54',
14:30:50-'NG-TC-W2K3-BASE_55',
14:30:50-'NG-TC-W2K3-BASE_56',
14:30:50-'NG-TC-W2K3-BASE_57',
14:30:50-'NG-TC-W2K3-BASE_58',
14:30:50-'NG-TC-W2K3-BASE_59',
14:30:50-'NG-TC-W2K3-BASE_6',
14:30:50-'NG-TC-W2K3-BASE_60',
14:30:50-'NG-TC-W2K3-BASE_61',
total = 65

Tree View Flat View (newer first) Flat View (older first)

Message #2151

as found in this thread : http://nsclient.org/nscp/discussion/message/2064#2064

there is a problem when the time of the client is greater than the server's time increasing the time by one minute resolves the problem. the latest release 0.3.8 and nightly build are affected

A quick fix would be to schedule an ntpdate followed by this command : date --set='+1 minutes'

i hope this problem will be corrected soon

Regards

Massimo

mtrento

04/29/11 13:18:24 (2 years ago)
Message #2111

Hello,

I'm facing the exact same problem. When I run more than 15 checks on some hosts, more than 30 on another, more than 100 on another, etc... Not every check seems to be received by NSCA...

I tested the ini with nsca commands like : NG-TC-W2K3-BASE*

Every time, the nsclient.log said that every check has been sent. Every time, NSCA said (in syslog) that only some checks have been received.

With wireshark, I have noted that not every check have been sent.

I tried this configuration with the official stable release (0.3.8), with some nightly build of nsclient (0.3.8-[...]-2316 for example), and with the latest nightly build of 0.3.9. I tried with, and withou chunk=10.

And I have never received all the check. Sometimes, I receive only 80 checks, sometime 150, sometime 190, etc...

If this bug has really been resolved, could you point me to the good version of NSClient ?

Thanks in advance.

misterdev

03/24/11 11:54:30 (2 years ago)
Message #1917

Hi Ryan,

do you still have problemes or did Michael's last change work for you? If you still had problems, I would test your config ...

Best regards,

Timo

timo

08/02/10 18:09:40 (3 years ago)
Message #1868

A fix for this issue in the latest nightly (out in a few minutes).

I found some issues with my socket closing code which might have caused a hard close of the socket (instead of a soft close).

Michael Medin

mickem

07/27/10 17:03:21 (3 years ago)
Message #1867
I ran it over the night here and had the following results:
```
wc -l /var/run/nagios/nagios.cmd
37611 /var/run/nagios/nagios.cmd
grep NG-TC-W2K3-BASE_99 /var/run/nagios/nagios.cmd |wc -l
189 /var/run/nagios/nagios.cmd
```
So at least for me "chunking" works...

I shall (now) start looking into why NSCA fails...

Michael Medin
mickem

07/27/10 08:52:03 (3 years ago)

Message #1865

Michael/Timo?.

Can either of you run this ini as a secondary ini and validate you receive all nsca messages (all 199)? I keep testing with a chuck=10 and it is sending all except the last ~10. Since it does them numeracially it is typically missing NG-TC-W2K3-BASE_90 through 99.

I am interested in your findings.

[modules]
[Settings]
[log]
[NSClient]
[NRPE]
[Check System]
[External Script]
[External Scripts]
[External Alias]
[NSCA Agent]
interval=120
[NSCA Commands]
NG-TC-W2K3-BASE_1=checkServiceState lanmanserver
NG-TC-W2K3-BASE_2=checkServiceState lanmanserver
NG-TC-W2K3-BASE_3=checkServiceState lanmanserver
NG-TC-W2K3-BASE_4=checkServiceState lanmanserver
NG-TC-W2K3-BASE_5=checkServiceState lanmanserver
NG-TC-W2K3-BASE_6=checkServiceState lanmanserver
NG-TC-W2K3-BASE_7=checkServiceState lanmanserver
NG-TC-W2K3-BASE_8=checkServiceState lanmanserver
NG-TC-W2K3-BASE_9=checkServiceState lanmanserver
NG-TC-W2K3-BASE_10=checkServiceState lanmanserver
NG-TC-W2K3-BASE_11=checkServiceState lanmanserver
NG-TC-W2K3-BASE_12=checkServiceState lanmanserver
NG-TC-W2K3-BASE_13=checkServiceState lanmanserver
NG-TC-W2K3-BASE_14=checkServiceState lanmanserver
NG-TC-W2K3-BASE_15=checkServiceState lanmanserver
NG-TC-W2K3-BASE_16=checkServiceState lanmanserver
NG-TC-W2K3-BASE_17=checkServiceState lanmanserver
NG-TC-W2K3-BASE_18=checkServiceState lanmanserver
NG-TC-W2K3-BASE_19=checkServiceState lanmanserver
NG-TC-W2K3-BASE_20=checkServiceState lanmanserver
NG-TC-W2K3-BASE_21=checkServiceState lanmanserver
NG-TC-W2K3-BASE_22=checkServiceState lanmanserver
NG-TC-W2K3-BASE_23=checkServiceState lanmanserver
NG-TC-W2K3-BASE_24=checkServiceState lanmanserver
NG-TC-W2K3-BASE_25=checkServiceState lanmanserver
NG-TC-W2K3-BASE_26=checkServiceState lanmanserver
NG-TC-W2K3-BASE_27=checkServiceState lanmanserver
NG-TC-W2K3-BASE_28=checkServiceState lanmanserver
NG-TC-W2K3-BASE_29=checkServiceState lanmanserver
NG-TC-W2K3-BASE_30=checkServiceState lanmanserver
NG-TC-W2K3-BASE_31=checkServiceState lanmanserver
NG-TC-W2K3-BASE_32=checkServiceState lanmanserver
NG-TC-W2K3-BASE_33=checkServiceState lanmanserver
NG-TC-W2K3-BASE_34=checkServiceState lanmanserver
NG-TC-W2K3-BASE_35=checkServiceState lanmanserver
NG-TC-W2K3-BASE_36=checkServiceState lanmanserver
NG-TC-W2K3-BASE_37=checkServiceState lanmanserver
NG-TC-W2K3-BASE_38=checkServiceState lanmanserver
NG-TC-W2K3-BASE_39=checkServiceState lanmanserver
NG-TC-W2K3-BASE_40=checkServiceState lanmanserver
NG-TC-W2K3-BASE_41=checkServiceState lanmanserver
NG-TC-W2K3-BASE_42=checkServiceState lanmanserver
NG-TC-W2K3-BASE_43=checkServiceState lanmanserver
NG-TC-W2K3-BASE_44=checkServiceState lanmanserver
NG-TC-W2K3-BASE_45=checkServiceState lanmanserver
NG-TC-W2K3-BASE_46=checkServiceState lanmanserver
NG-TC-W2K3-BASE_47=checkServiceState lanmanserver
NG-TC-W2K3-BASE_48=checkServiceState lanmanserver
NG-TC-W2K3-BASE_49=checkServiceState lanmanserver
NG-TC-W2K3-BASE_50=checkServiceState lanmanserver
NG-TC-W2K3-BASE_51=checkServiceState lanmanserver
NG-TC-W2K3-BASE_52=checkServiceState lanmanserver
NG-TC-W2K3-BASE_53=checkServiceState lanmanserver
NG-TC-W2K3-BASE_54=checkServiceState lanmanserver
NG-TC-W2K3-BASE_55=checkServiceState lanmanserver
NG-TC-W2K3-BASE_56=checkServiceState lanmanserver
NG-TC-W2K3-BASE_57=checkServiceState lanmanserver
NG-TC-W2K3-BASE_58=checkServiceState lanmanserver
NG-TC-W2K3-BASE_59=checkServiceState lanmanserver
NG-TC-W2K3-BASE_60=checkServiceState lanmanserver
NG-TC-W2K3-BASE_61=checkServiceState lanmanserver
NG-TC-W2K3-BASE_62=checkServiceState lanmanserver
NG-TC-W2K3-BASE_63=checkServiceState lanmanserver
NG-TC-W2K3-BASE_64=checkServiceState lanmanserver
NG-TC-W2K3-BASE_65=checkServiceState lanmanserver
NG-TC-W2K3-BASE_66=checkServiceState lanmanserver
NG-TC-W2K3-BASE_67=checkServiceState lanmanserver
NG-TC-W2K3-BASE_68=checkServiceState lanmanserver
NG-TC-W2K3-BASE_69=checkServiceState lanmanserver
NG-TC-W2K3-BASE_70=checkServiceState lanmanserver
NG-TC-W2K3-BASE_71=checkServiceState lanmanserver
NG-TC-W2K3-BASE_72=checkServiceState lanmanserver
NG-TC-W2K3-BASE_73=checkServiceState lanmanserver
NG-TC-W2K3-BASE_74=checkServiceState lanmanserver
NG-TC-W2K3-BASE_75=checkServiceState lanmanserver
NG-TC-W2K3-BASE_76=checkServiceState lanmanserver
NG-TC-W2K3-BASE_77=checkServiceState lanmanserver
NG-TC-W2K3-BASE_78=checkServiceState lanmanserver
NG-TC-W2K3-BASE_79=checkServiceState lanmanserver
NG-TC-W2K3-BASE_80=checkServiceState lanmanserver
NG-TC-W2K3-BASE_81=checkServiceState lanmanserver
NG-TC-W2K3-BASE_82=checkServiceState lanmanserver
NG-TC-W2K3-BASE_83=checkServiceState lanmanserver
NG-TC-W2K3-BASE_84=checkServiceState lanmanserver
NG-TC-W2K3-BASE_85=checkServiceState lanmanserver
NG-TC-W2K3-BASE_86=checkServiceState lanmanserver
NG-TC-W2K3-BASE_87=checkServiceState lanmanserver
NG-TC-W2K3-BASE_88=checkServiceState lanmanserver
NG-TC-W2K3-BASE_89=checkServiceState lanmanserver
NG-TC-W2K3-BASE_90=checkServiceState lanmanserver
NG-TC-W2K3-BASE_91=checkServiceState lanmanserver
NG-TC-W2K3-BASE_92=checkServiceState lanmanserver
NG-TC-W2K3-BASE_93=checkServiceState lanmanserver
NG-TC-W2K3-BASE_94=checkServiceState lanmanserver
NG-TC-W2K3-BASE_95=checkServiceState lanmanserver
NG-TC-W2K3-BASE_96=checkServiceState lanmanserver
NG-TC-W2K3-BASE_97=checkServiceState lanmanserver
NG-TC-W2K3-BASE_98=checkServiceState lanmanserver
NG-TC-W2K3-BASE_99=checkServiceState lanmanserver
NG-TC-W2K3-BASE_100=checkServiceState lanmanserver
NG-TC-W2K3-BASE_101=checkServiceState lanmanserver
NG-TC-W2K3-BASE_102=checkServiceState lanmanserver
NG-TC-W2K3-BASE_103=checkServiceState lanmanserver
NG-TC-W2K3-BASE_104=checkServiceState lanmanserver
NG-TC-W2K3-BASE_105=checkServiceState lanmanserver
NG-TC-W2K3-BASE_106=checkServiceState lanmanserver
NG-TC-W2K3-BASE_107=checkServiceState lanmanserver
NG-TC-W2K3-BASE_108=checkServiceState lanmanserver
NG-TC-W2K3-BASE_109=checkServiceState lanmanserver
NG-TC-W2K3-BASE_110=checkServiceState lanmanserver
NG-TC-W2K3-BASE_111=checkServiceState lanmanserver
NG-TC-W2K3-BASE_112=checkServiceState lanmanserver
NG-TC-W2K3-BASE_113=checkServiceState lanmanserver
NG-TC-W2K3-BASE_114=checkServiceState lanmanserver
NG-TC-W2K3-BASE_115=checkServiceState lanmanserver
NG-TC-W2K3-BASE_116=checkServiceState lanmanserver
NG-TC-W2K3-BASE_117=checkServiceState lanmanserver
NG-TC-W2K3-BASE_118=checkServiceState lanmanserver
NG-TC-W2K3-BASE_119=checkServiceState lanmanserver
NG-TC-W2K3-BASE_120=checkServiceState lanmanserver
NG-TC-W2K3-BASE_121=checkServiceState lanmanserver
NG-TC-W2K3-BASE_122=checkServiceState lanmanserver
NG-TC-W2K3-BASE_123=checkServiceState lanmanserver
NG-TC-W2K3-BASE_124=checkServiceState lanmanserver
NG-TC-W2K3-BASE_125=checkServiceState lanmanserver
NG-TC-W2K3-BASE_126=checkServiceState lanmanserver
NG-TC-W2K3-BASE_127=checkServiceState lanmanserver
NG-TC-W2K3-BASE_128=checkServiceState lanmanserver
NG-TC-W2K3-BASE_129=checkServiceState lanmanserver
NG-TC-W2K3-BASE_130=checkServiceState lanmanserver
NG-TC-W2K3-BASE_131=checkServiceState lanmanserver
NG-TC-W2K3-BASE_132=checkServiceState lanmanserver
NG-TC-W2K3-BASE_133=checkServiceState lanmanserver
NG-TC-W2K3-BASE_134=checkServiceState lanmanserver
NG-TC-W2K3-BASE_135=checkServiceState lanmanserver
NG-TC-W2K3-BASE_136=checkServiceState lanmanserver
NG-TC-W2K3-BASE_137=checkServiceState lanmanserver
NG-TC-W2K3-BASE_138=checkServiceState lanmanserver
NG-TC-W2K3-BASE_139=checkServiceState lanmanserver
NG-TC-W2K3-BASE_140=checkServiceState lanmanserver
NG-TC-W2K3-BASE_141=checkServiceState lanmanserver
NG-TC-W2K3-BASE_142=checkServiceState lanmanserver
NG-TC-W2K3-BASE_143=checkServiceState lanmanserver
NG-TC-W2K3-BASE_144=checkServiceState lanmanserver
NG-TC-W2K3-BASE_145=checkServiceState lanmanserver
NG-TC-W2K3-BASE_146=checkServiceState lanmanserver
NG-TC-W2K3-BASE_147=checkServiceState lanmanserver
NG-TC-W2K3-BASE_148=checkServiceState lanmanserver
NG-TC-W2K3-BASE_149=checkServiceState lanmanserver
NG-TC-W2K3-BASE_150=checkServiceState lanmanserver
NG-TC-W2K3-BASE_151=checkServiceState lanmanserver
NG-TC-W2K3-BASE_152=checkServiceState lanmanserver
NG-TC-W2K3-BASE_153=checkServiceState lanmanserver
NG-TC-W2K3-BASE_154=checkServiceState lanmanserver
NG-TC-W2K3-BASE_155=checkServiceState lanmanserver
NG-TC-W2K3-BASE_156=checkServiceState lanmanserver
NG-TC-W2K3-BASE_157=checkServiceState lanmanserver
NG-TC-W2K3-BASE_158=checkServiceState lanmanserver
NG-TC-W2K3-BASE_159=checkServiceState lanmanserver
NG-TC-W2K3-BASE_160=checkServiceState lanmanserver
NG-TC-W2K3-BASE_161=checkServiceState lanmanserver
NG-TC-W2K3-BASE_162=checkServiceState lanmanserver
NG-TC-W2K3-BASE_163=checkServiceState lanmanserver
NG-TC-W2K3-BASE_164=checkServiceState lanmanserver
NG-TC-W2K3-BASE_165=checkServiceState lanmanserver
NG-TC-W2K3-BASE_166=checkServiceState lanmanserver
NG-TC-W2K3-BASE_167=checkServiceState lanmanserver
NG-TC-W2K3-BASE_168=checkServiceState lanmanserver
NG-TC-W2K3-BASE_169=checkServiceState lanmanserver
NG-TC-W2K3-BASE_170=checkServiceState lanmanserver
NG-TC-W2K3-BASE_171=checkServiceState lanmanserver
NG-TC-W2K3-BASE_172=checkServiceState lanmanserver
NG-TC-W2K3-BASE_173=checkServiceState lanmanserver
NG-TC-W2K3-BASE_174=checkServiceState lanmanserver
NG-TC-W2K3-BASE_175=checkServiceState lanmanserver
NG-TC-W2K3-BASE_176=checkServiceState lanmanserver
NG-TC-W2K3-BASE_177=checkServiceState lanmanserver
NG-TC-W2K3-BASE_178=checkServiceState lanmanserver
NG-TC-W2K3-BASE_179=checkServiceState lanmanserver
NG-TC-W2K3-BASE_180=checkServiceState lanmanserver
NG-TC-W2K3-BASE_181=checkServiceState lanmanserver
NG-TC-W2K3-BASE_182=checkServiceState lanmanserver
NG-TC-W2K3-BASE_183=checkServiceState lanmanserver
NG-TC-W2K3-BASE_184=checkServiceState lanmanserver
NG-TC-W2K3-BASE_185=checkServiceState lanmanserver
NG-TC-W2K3-BASE_186=checkServiceState lanmanserver
NG-TC-W2K3-BASE_187=checkServiceState lanmanserver
NG-TC-W2K3-BASE_188=checkServiceState lanmanserver
NG-TC-W2K3-BASE_189=checkServiceState lanmanserver
NG-TC-W2K3-BASE_190=checkServiceState lanmanserver
NG-TC-W2K3-BASE_191=checkServiceState lanmanserver
NG-TC-W2K3-BASE_192=checkServiceState lanmanserver
NG-TC-W2K3-BASE_193=checkServiceState lanmanserver
NG-TC-W2K3-BASE_194=checkServiceState lanmanserver
NG-TC-W2K3-BASE_195=checkServiceState lanmanserver
NG-TC-W2K3-BASE_196=checkServiceState lanmanserver
NG-TC-W2K3-BASE_197=checkServiceState lanmanserver
NG-TC-W2K3-BASE_198=checkServiceState lanmanserver
NG-TC-W2K3-BASE_199=checkServiceState lanmanserver
;[LUA Scripts]

Message #1864

Timo, Can you give me some details on what your test scenario(s) were? What are the client versions (windows 2003/2008/xp?) What is interval? How many nsca responses are you forwarding off per interval?

THanks. I am hoping this is just something I have overlooked.

tavor999

07/20/10 22:06:38 (3 years ago)
Message #1863

Hi Michael,

we have tested the latest nightly and it seems to work as expected. No problems at all anymore (at least in our setting; I hope Ryan gets his problems sorted out as well) ...

It would be great if you kept the chunk option in the next major release.

Thanks a lot,

Timo

timo

07/20/10 17:18:00 (3 years ago)
Message #1862

Humm...

Maybe it is time to replace NSCA then?

(Or spend some time on the NSCA server trying to see why it fails)

I have to say for the next 5 days I wont have much time to tinker with this but I can help you again after that...

But I am leaning towards replacing NSCA since NSCA is a pretty crappy protocol, but I am not sure you ind it to be a viable option?

(another one would be to parallelize the calls, I guess)

Michael Medin

mickem

07/20/10 07:05:06 (3 years ago)
Message #1861

I did a little testing on the nightly build and here is what I have found.

I can see the chunking in nsca messages roll in on the server side. Of my 300 test service checks the client again appears to be running all tests without an error. I have tried chunk from 50-10. At 50 I saw about 150 messages per interval (60 sec while testing). At 10sec chunk I saw 290 messages. The problem is with a 10 sec chunk the client is now taking ~40sec to complete that cycle. The tests are taking ~6sec to run, then NSCA takes the remainder of the time opening and closing sessions to send that information. I will continue testing but I don't think this solution is going to work. It appears we will have to lower the chunk so low that it effectively causes too much overhead with session initiation, etc.

NOTE: I had hoped to utilize the "report" feature to tell our clients to only report warning,critica,unknown. The problem is that some of our nsca test will have performance information we need to get back to PNP for storage. So I cannot blanketly shut off the messages. If I could some how apply that "report" feature per test, or test I think my number of tests reporting per interval would be manageable.

Michael...thx for continued help. Hopefully our engagement will proceed shortly :)

Ryan Ash

tavor999

07/19/10 16:40:04 (3 years ago)
Message #1860

Sorry for that... try now... I have actually tested this (almost)

Michael Medin

mickem

07/19/10 09:53:02 (3 years ago)
Message #1859

Good idea timo. That would probably work because we didn't notice this issue around ~50 service checks. Testing at 150 and beyond we starting seeing this issue.

tavor999

07/16/10 14:48:48 (3 years ago)
Message #1858

Wow, thanks a lot for your extremely fast reply and for your help.

Unfortunately, the new option does not seem to work. If I execute "nsclient++ /test" it shows that all checks are executed and sent afterwards. It does not seem to split the results into chunks ...

It would be great if you could check that again ...

Best regards and thanks for your support,

Timo

timo

07/16/10 14:47:27 (3 years ago)
Message #1857

I confirmed nothing in the syslog is indicating an error with the nsca transmission. I also did a network capture and both sides to confirm the NSCA message are not leaving the client. Reminder of the 300 tests per interval we usually get something like 50 responses. Sometimes more, sometimes less...they are not the same 50 every time...

tavor999

07/16/10 14:46:57 (3 years ago)
Message #1856
Try the latest nightly.

It adds a new option like so:
```
[NSCA Agent]
...
chunk=10
```
Which , in theory, will send only 10 results each time.

Notice I haven't had time to test this at all: so do this ON A DEVELOPMENT BOX!!'''

Michael Medin
mickem

07/16/10 07:48:29 (3 years ago)
Message #1855

Hi Michael,

as tavor999 and many others, we are having similar problems. We tried recompiling nsca client and server with different buffer sizes but nothing helped. I believe it would be of great value if you added an option in the NSC.ini (and the necessary code in the backend) to set the number of check results that are sent in one block to the nsca server.

For example:

If an option "max_checks_sent=5" is set, nsclient++ only executes the first five checks and sends the results right away. After sending the results, the next five checks are executed and sent. After all checks are executed, nsclient++ starts at the beginning again.

Another possible solution would be to execute all checks but to send them in blocks. If "max_checks_sent=5", nsclient++ could execute all checks and store the results in a temporary file. Afterwards nsca could send them in blocks of 5.

Without any solution, we would not be able to use nagios at all. Active checks are not an option. Right now, we execute many instances of nsclient++ on the same machine to realize the behaviour described above. That is pretty ugly and it would be great if you could help ...

Best regards and thanks for your work,

Timo

timo

07/16/10 02:07:52 (3 years ago)
Message #1854

Right..

And there is nothin in the syslog?

NSCA server does not support error reporting to the "client" so often a good place to find errors can be there.

But mainly such errors are related to "time" and/or encryption so probably doubtful.

Anyways, I shall setup some stress test environments this weekend and see if I can get similar results.

Michael Medin

mickem

07/15/10 22:04:46 (3 years ago)

Message #1853

Thanks for your response.

Originally I found with with a combination of different ini files with lots of eventlog checks. To simplify my troubleshooting I made a generic INI file 1-300 checks using checkService.

NG-TC-W2K3-BASE_1=checkServiceState lanmanserver NG-TC-W2K3-BASE_2=checkServiceState lanmanserver NG-TC-W2K3-BASE_3=checkServiceState lanmanserver NG-TC-W2K3-BASE_4=checkServiceState lanmanserver NG-TC-W2K3-BASE_5=checkServiceState lanmanserver NG-TC-W2K3-BASE_6=checkServiceState lanmanserver NG-TC-W2K3-BASE_7=checkServiceState lanmanserver NG-TC-W2K3-BASE_8=checkServiceState lanmanserver NG-TC-W2K3-BASE_9=checkServiceState lanmanserver NG-TC-W2K3-BASE_10=checkServiceState lanmanserver NG-TC-W2K3-BASE_11=checkServiceState lanmanserver NG-TC-W2K3-BASE_12=checkServiceState lanmanserver NG-TC-W2K3-BASE_13=checkServiceState lanmanserver

2010-07-13 12:35:37: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_1
2010-07-13 12:35:37: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:37: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:37: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:37: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_10
2010-07-13 12:35:37: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:37: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:37: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_100
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_101
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_102
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_103
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_104
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_105
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_106
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_107
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_108
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_109
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_11
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_110
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''
2010-07-13 12:35:38: debug:modules\NSCAAgent\NSCAThread.cpp:205: Executing (from NSCA): NG-TC-W2K3-BASE_111
2010-07-13 12:35:38: debug:NSClient++.cpp:1106: Injecting: checkServiceState: lanmanserver
2010-07-13 12:35:38: debug:NSClient++.cpp:1142: Injected Result: OK 'OK: All services are in their appropriate state.'
2010-07-13 12:35:38: debug:NSClient++.cpp:1143: Injected Performance Result: ''

The nsclient.log on the client will show ALL tests were run and looked good. Just when I look for the corresponding NSCA message it is not received (or sent, check packet capture).

The message output is small "All services are in their appropriate state".

Thanks again...

Message #1851

What kind of checks?

There is a fairly important (known) "issue" with NSCA and that is the payload length if you "overflow" it it will "bail out" or ignore the entry (don't recall which off the top of my head) so NSCA can be difficult.

If this is a show stopper I could probably work something out (but then I would need to know the exact nature of the issue, ie. simulate it here).

Michael Medin

mickem

07/13/10 21:39:28 (3 years ago)

Tree View Flat View (newer first) Flat View (older first)

Subscriptions

Download in other formats:

RSS Feed