October 03, 2022, 01:27:40 PM

Author Topic: Update (issue fixed in build 2284)  (Read 7193 times)

MrPhelps

  • Member
  • *
  • Posts: 2
Update (issue fixed in build 2284)
« on: August 29, 2005, 05:39:46 PM »
I am having trouble using discover.  It seams to hang when it get\'s to the last few pings waiting for a reply.  I can not stop the scan from the interface.  If I kill NEWTScanner manually, it always restarts. Scanning about 300 objects.

NEWTPro v2.0 build 2274, running on XP sp1.
Thax,

Komodo Support

  • Administrator
  • Member
  • *****
  • Posts: 2701
  • Dayton, Ohio, USA
    • Komodo Laboratories LLC
Update (issue fixed in build 2284)
« Reply #1 on: August 29, 2005, 06:54:12 PM »
Hello.  When you\'re discovering and you have a few pings or threads left, are you able to see if any machines say \"Scanning OS...\" under the Operating System column of the Primary tab?  And may I ask how long you\'re waiting?  It can take up to a minute for these last few threads to finish.

We are working on a user-definable timeout setting for scanning the OS, but this will take some time.  You could also try turning this feature off before discoverying to make sure this is the culprit.

Since the NEWTScanner isn\'t used during a discovery, killing it shouldn\'t affect anything.

MrPhelps

  • Member
  • *
  • Posts: 2
Update (issue fixed in build 2284)
« Reply #2 on: August 31, 2005, 09:51:29 AM »
Thanks for the reply.  I did turn off the OS scan on the Discover, but that seems to have no effect.  All of the Scan Status columns say \"Hostname Resolved\".  The message in the status line says \"Pings sent, waiting for replies (5 threads left)\".  I let it run overnight.

Komodo Support

  • Administrator
  • Member
  • *****
  • Posts: 2701
  • Dayton, Ohio, USA
    • Komodo Laboratories LLC
Update (issue fixed in build 2284)
« Reply #3 on: September 01, 2005, 02:50:42 AM »
We apologize you\'re having difficulty with NEWT Pro.  We haven\'t yet heard of a problem like this, but we will contact you through email to see if we can nail down this issue.

jmcintyre

  • Member
  • *
  • Posts: 2
Update (issue fixed in build 2284)
« Reply #4 on: October 31, 2005, 05:01:05 PM »
I just ran a discover for over 37,000 ip addresses, and the software is now hung.  The NEWTPro.exe process is running at 99% CPU.  The status message at the bottom says: \"Pings sent, waiting for replies (0 threads left).  The middle message says: \"maching name (Windows NT 4.0) at ip address (87%).   The other data areas on the bottom say: Estimated time left 03m10s,  Elapsed time 48m30s,  37329 objects, 0 objects selected.    The times are not changing.  It has been sitting this way for 30 minutes.  The Task manager shows memory usage increasing for the NEWTPro.exe process.  

I was able to do this same scan earlier in the day and it worked fine.  It seems to be stuck on one machine.  I can ping that machine from the PC running NEWT.  Is there any way to get it to respond?  Hitting the stop button does nothing.

Komodo Support

  • Administrator
  • Member
  • *****
  • Posts: 2701
  • Dayton, Ohio, USA
    • Komodo Laboratories LLC
Update (issue fixed in build 2284)
« Reply #5 on: October 31, 2005, 08:26:05 PM »
We believe we\'ve recreated this issue.  It seems to only happen on very large IP ranges.  We are working on the problem now and should have a fix for you very soon.

We will contact you by private email if that\'s ok.

jmcintyre

  • Member
  • *
  • Posts: 2
Update (issue fixed in build 2284)
« Reply #6 on: November 01, 2005, 08:54:24 AM »
I left NEWT running through the night.  When I came in this morning the scan was finished.   The status messages at the bottom say: IP discovery took 01h49m17s, 5309 objects.
I will wait for your email.

Komodo Support

  • Administrator
  • Member
  • *****
  • Posts: 2701
  • Dayton, Ohio, USA
    • Komodo Laboratories LLC
Update (issue fixed in build 2284)
« Reply #7 on: November 01, 2005, 12:58:52 PM »
It sounds like you\'re not running into the same problem we were during our test.  Our test never finished because NEWT wasn\'t limiting its thread count to the safe range of about 1,000.  But in a true LAN, or in your case, NEWT doesn\'t seem to go over the thread limit because machines return data more quickly.  We believe your issue is during the \"cleanup\" where NEWT removes dead IPs and NEWT isn\'t displaying a \"cleanup status\".

Unfortunately there is nothing we can do immediately about the time it takes to clean up, but we could show the user the cleanup is happening.  We will make this particular change soon in a maintenance.  Most users do not discover such large ranges so the cleanup time is so fast there doesn\'t need to be a status shown, but we are working to speed up the IP Discovery cleanup, IP Discovery itself and other parts of the software dramatically in the future.

Komodo Support

  • Administrator
  • Member
  • *****
  • Posts: 2701
  • Dayton, Ohio, USA
    • Komodo Laboratories LLC
Update (issue fixed in build 2284)
« Reply #8 on: December 23, 2005, 04:50:17 PM »
We\'re now happy to announce this issue should be fixed in build 2284, which is available on our main home page.