HM freezes after update from V8.86 to V9.06

When you post information about some problem, please include the following details: - OS version (e.g. Windows 2000 Professional SP3); HostMonitor version; problem description.
Post Reply
apaitoperations
Posts: 40
Joined: Thu Feb 24, 2011 1:55 am

Status Checking

Post by apaitoperations »

Hello Alex,

We use Views to display only relevant testitems.

In View-Properties there is NO Option to filter for Status "Checking" - so how can i see these Testitems ?

I will check total thread-count again when HM stops working next time

wbr
Georg Höllebauer
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

In View-Properties there is NO Option to filter for Status "Checking" - so how can i see these Testitems ?
You cannot use Views for this purpose. Only Folders.
If new debuglog1.txt does not show any test items, then skip this.
H'm... if Windows Task Manager will show much less threads, this means HostMonitor "sees" problem that does not exist :roll:

Regards
Alex
apaitoperations
Posts: 40
Joined: Thu Feb 24, 2011 1:55 am

Post by apaitoperations »

no - new debuglog doesnt show any such items :-(

I will compare thread-count debuglog <-> Windows-Task-manager next time.

I have to many testitems to display them all -> RCC will stop responding or become so slow if i display all testitems :-(

wbr
Georg Höllebauer
rc
Posts: 100
Joined: Mon Aug 01, 2005 7:51 am

Post by rc »

Hi Alex,

I have in our 2nd HostMonitor installation already recorded the number of threads. Below the values ​​from the debug log and threads during the last freezing:

Debuglog:
---- during freezing ----
12.11.2012 11:46:48
Timer1: 1 12.11.2012 11:46:47
Timer2: 1
PoolRecAvail: 4096
TTLimit1: 64
TCnt1: 12952
TTThreads: 1001
LIdx: 6162
ATCnt2: 32

---- during freezing ----
12.11.2012 11:47:26
Timer1: 1 12.11.2012 11:47:25
Timer2: 1
PoolRecAvail: 4096
TTLimit1: 64
TCnt1: 12952
TTThreads: 1001
LIdx: 6162
ATCnt2: 32

---- after restart ----
12.11.2012 12:04:17
Timer1: 1 12.11.2012 12:04:17
Timer2: 1
PoolRecAvail: 4096
TTLimit1: 64
TCnt1: 12952
TTThreads: 16
LIdx: 12952
ATCnt2: 32


thread log during and after freezing:
[12.11.2012 11:08:44] PerfCounter - Prozess(hostmon)\Threads Ok 55
[12.11.2012 11:13:49] PerfCounter - Prozess(hostmon)\Threads Ok 41
[12.11.2012 11:18:53] PerfCounter - Prozess(hostmon)\Threads Ok 40
[12.11.2012 11:23:57] PerfCounter - Prozess(hostmon)\Threads Ok 22
[12.11.2012 11:29:01] PerfCounter - Prozess(hostmon)\Threads Ok 25
[12.11.2012 11:34:05] PerfCounter - Prozess(hostmon)\Threads Ok 23
[12.11.2012 11:39:08] PerfCounter - Prozess(hostmon)\Threads Ok 21
[12.11.2012 11:44:11] PerfCounter - Prozess(hostmon)\Threads Ok 21
[12.11.2012 11:49:14] PerfCounter - Prozess(hostmon)\Threads Ok 21
[12.11.2012 11:54:18] PerfCounter - Prozess(hostmon)\Threads Ok 21
[12.11.2012 11:59:27] PerfCounter - Prozess(hostmon)\Threads Ok 20
[12.11.2012 12:04:33] PerfCounter - Prozess(hostmon)\Threads Ok 38
[12.11.2012 12:09:37] PerfCounter - Prozess(hostmon)\Threads Ok 51
[12.11.2012 12:14:44] PerfCounter - Prozess(hostmon)\Threads Ok 21
[12.11.2012 12:19:48] PerfCounter - Prozess(hostmon)\Threads Ok 67
[12.11.2012 12:24:53] PerfCounter - Prozess(hostmon)\Threads Ok 32
[12.11.2012 12:29:58] PerfCounter - Prozess(hostmon)\Threads Ok 23
[12.11.2012 12:35:04] PerfCounter - Prozess(hostmon)\Threads Ok 23


I note: the values ​​for TTThreads and hostmon\Threads are extremely interdependent.

And now?

Regards
Enrico
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

I note: the values ​​for TTThreads and hostmon\Threads are extremely interdependent.
Could you please explain what exactly this means? How many threads was used by hostmon.exe process according to Task Manager or Performance Counters (while TTThreads: 1001)?

Regards
Alex
rc
Posts: 100
Joined: Mon Aug 01, 2005 7:51 am

Post by rc »

I loged with 2nd hostmonitor threads was used by 1st hostmon.exe process according to Performance Counters. Is there any difference to task manager? I mean this is the same value. I have compare the values live and the values are identically.

Regards Enrico
rc
Posts: 100
Joined: Mon Aug 01, 2005 7:51 am

Post by rc »

By the way: Here are current values from debug log:

22.11.2012 15:23:50
Timer1: 1 22.11.2012 15:23:49
Timer2: 1
PoolRecAvail: 4096
TTLimit1: 64
TCnt1: 12961
TTThreads: 708
LIdx: 12961
ATCnt2: 32
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

Then I do not understand anything :(
1000 threads, 1000 started tests but just 18 items in checking status for more than 40 sec?
rc
Posts: 100
Joined: Mon Aug 01, 2005 7:51 am

Post by rc »

I interpret my debuglog so that the value for TTThreads increases continuously up to 1000 (in duration of 4 days), and then after a short time, the system no longer performs tests. At this time all other values​​ such as threads, handles and cpu usage are normal.

So the question is: Why is the value for TTThreads constantly increasing after restart?

The test frequency is now standard to me all the time at 37 tests / sec.
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

So the question is: Why is the value for TTThreads constantly increasing after restart?
That's very easy to answer - because some test threads hung.
Hard questions:
1) why it hungs
2) if test threads hung, why you do not see "Unknown"/"Timed Out" and "Checking" test items? Its just impossible!!!!

Regards
Alex
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

BTW: is this complete list?
-----------
thread log during and after freezing:
[12.11.2012 11:08:44] PerfCounter - Prozess(hostmon)\Threads Ok 55
[12.11.2012 11:13:49] PerfCounter - Prozess(hostmon)\Threads Ok 41
......
----------
or you just copied 18 items while there are much more in the log?

Regards
Alex
rc
Posts: 100
Joined: Mon Aug 01, 2005 7:51 am

Post by rc »

Alex, you got me completely misunderstood.
1000 threads, 1000 started tests but just 18 items in checking status for more than 40 sec?
The 18 lines are not from debuglog. This is a cutout from another logfile from the other Hostmonitor, that monitors continuously the main host.

Enrico
apaitoperations
Posts: 40
Joined: Thu Feb 24, 2011 1:55 am

comparision Threads

Post by apaitoperations »

HM ist frozen again and in debuglog there is again 1001 :
----
2012-11-22 17:11:25
Timer1: 1 2012-11-22 17:11:24
Timer2: 1
PoolRecAvail: 4096
TTLimit1: 128
TCnt1: 9718
TTThreads: 1001
LIdx: 3858
ATCnt2: 32

But with Taskmanager i see that Hostmon.exe has only 19 threads

wbr
Georg Höllebauer
KS-Soft
Posts: 13012
Joined: Wed Apr 03, 2002 6:00 pm
Location: USA
Contact:

Post by KS-Soft »

The 18 lines are not from debuglog. This is a cutout from another logfile from the other Hostmonitor, that monitors continuously the main host
Then I don't understand why "TTThreads and hostmon\Threads are extremely interdependent"?

Ok, so there are no hung threads, no "checking" and "timed out" items?
And hostmon.exe process always has < 100 threads (even when monitoring is stopped and TTThreads>1000)?

Then this is completely different issue - there is no real problem while HostMonitor "sees" the problem because of some mistake related to thread counter.
Ok, we will check everything related to this counter and may be provide some quick fix with option that disables this counter check...

Regards
Alex
apaitoperations
Posts: 40
Joined: Thu Feb 24, 2011 1:55 am

thread counter in Hostmonitor

Post by apaitoperations »

Alex,

So there is an internal check within Hostmonitor to count its threads ?

And it has a limit ?

wbr
Georg
Post Reply