Network Automation Software We are currently migrating this forum
over to our HelpSystems domain. Please
post all new threads in our new
HelpSystems Community Portal.
Post to the HelpSystems Forum
You are not currently logged on. You must be logged on in order to post. Log on
Or Create a new account
AutoMate Discussion
Decrease font size
Increase font size
Topic Title: DNS change in an enterprise environment requires an Automate service restart to update
Topic Summary: W2K8 R2, Using 8.0.9.0., detect and use new DNS
Created On: 06/28/2013 01:23 PM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 06/28/2013 01:23 PM
User is offline View Users Profile Print this message

Author Icon
DLJNA
Apprentice (20-49)

Posts: 29
Joined: 11/17/2005

When old DNS servers go offline and new DNS servers brought online, Automate is not using the new DNS servers (in other words, monitored shares requiring DNS resolution are no longer processed). The File Shares being accessed are clustered, and we don't use static IPs. The server however seems to detect the new DNS when interactively tested at console (not using Automate).

Is there a best practice for handling this (I'm just looking for high level guidance)?

Is there something I should change in the server, or trap the error and somehow force Automate service to redetect the new servers quickly without a restart? Or code to re-resolve the DNS every single time a task is run (many hundreds of times per day)?

Thanks!

 Category Survey
 07/01/2013 10:21 AM
User is offline View Users Profile Print this message

Author Icon
Liz Casale
AUTOMATE EMPLOYEE

Posts: 940
Joined: 05/23/2011

Hi DLJNA,

Are you using full UNC paths to reference these locations or IP addresses currently?

 Category Survey


-------------------------
Liz Casale
Technical Support Representative
(213) 738-6966
Network Automation, Inc.

--------------------------------------------------------------------------------

AM5 5.5.5.1; AM6 6.2.8.0; AM7 7.1.3.0; AM8 8.0.9.0; AM9 9.0.3.4
BPA7 7.1.3.0; BPA8 8.0.9.0; BPA9 9.0.3.4
 07/01/2013 11:50 AM
User is offline View Users Profile Print this message

Author Icon
DLJNA
Apprentice (20-49)

Posts: 29
Joined: 11/17/2005

Thanks Liz.

Using UNC paths, as the fileshares are clustered. It's worked fine for quite a while, but they need to rotate their DNS servers periodically, and when they take the old ones offline, the new ones aren't seen by Automate until the Automate services are restarted. The operating system sees the changes right away.

Just looking for high-level best practices in what might be a good approach to accommodate this kind of scenario.

Much appreciated!

 Category Survey
 07/02/2013 04:37 PM
User is offline View Users Profile Print this message

Author Icon
Liz Casale
AUTOMATE EMPLOYEE

Posts: 940
Joined: 05/23/2011

Hi DLJA,

Thank you for your reply. The File Watch trigger relies on the DNS information from Windows. After the DNS comes back online, if you browse to one of the File Watch locations outside of AutoMate, is it accessible immediately? The DNS may be being cached in Windows. If you do a DNS Flush (ipconfig -flushdns) via command prompt after the DNS is resolvable again, is there any change in the behavior of AutoMate?

 Category Survey


-------------------------
Liz Casale
Technical Support Representative
(213) 738-6966
Network Automation, Inc.

--------------------------------------------------------------------------------

AM5 5.5.5.1; AM6 6.2.8.0; AM7 7.1.3.0; AM8 8.0.9.0; AM9 9.0.3.4
BPA7 7.1.3.0; BPA8 8.0.9.0; BPA9 9.0.3.4
 07/03/2013 06:40 AM
User is offline View Users Profile Print this message

Author Icon
DLJNA
Apprentice (20-49)

Posts: 29
Joined: 11/17/2005

Thank you again Liz.

I'm not onsite, but I'm pretty sure that would also have fixed the situation (rather than overkill with a restart).

In an enterprise environment, how would Network Automation recommend addressing this scenario in its tasks?

Build an error condition that routes to a label and runs a DNS flush, waits for a few seconds, and then returns to retry the task?

There are dozens of network calls during a given process run, and potentially 10 of these processes that can run concurrently at any time. And there are multiple servers all running the same Automate services running the same Tasks at the same time. Are there any 'gotchas' as far as running 10 instances of the same task (on just one server), where one of those encounters a network timeout and does a DNS flush and resumes, where it will impact the other concurrent tasks that have not yet encountered that condition yet?

In other words, does Automate resolve DNS addresses at start time, and store those during the 'lifetime' of a task's run, or does it dynamically update any network calls according to each individual Automate Task function?

As an example: Let's say a Task (instance A) copies some files from a network share, does some local processing on those files (taking some time), and then puts the documents back to the remote share. What if, between those two network accesses, another instance of the same Task (instance B) had already encountered that DNS resolution error, and done a DNS flush? Would the Task (instance A) use the updated DNS resolution? Would the remaining steps of other Task (instance B) use the updated resolution too?

Apologies for the rather nasty question, but I need my processes to accommodate this environment. Unfortunately, it won't be acceptable for me to propose taking the servers offline and doing a DNS flush - I need to factor that changing condition into the automation.

Thanks so much again!

 Category Survey
 07/03/2013 09:03 AM
User is offline View Users Profile Print this message

Author Icon
Hamid Noorbakhsh
HelpSystems

Posts: 454
Joined: 07/29/2008

DLJNA,
I am sorry to jump in here, but as far as I know our products do not cache any DNS entries. We use Windows DNS lookup and all this is handled through Windows itself (whether to resolve the entry again or use a cache version), so it all depends on your machine.

I THINK if you reduce your cache lifespan (even to zero <- will do a look up every time, it will be accurate but slower) this might solve your situation. (Of course I have not tried this but seems reasonable). Please look here and here. It might give you ideas and might help you resolve your issue easier than what you explained. Worst case scenario is that you will have to 'flushdns' on error and this in fact is good for the remainder of the tasks running since they need the new information (if some dns is not changed, oh well its the same address and it will do a look up of the same again)

I hope this will help you to reduce your failed tasks and allow Automate to run smoothly.

Please feel free to discuss this further if this does or does not resolve your issue since we will be curious to know how you get this working the way you want as I am sure someone else might use this information later.

 Category Survey


-------------------------
Thank you for choosing AutoMate

Hamid Noorbakhsh | Senior Software Engineer
AutoMate | A Division of HelpSystems
HelpSystems, LLC | T: +1.213.738.1700
 07/03/2013 04:12 PM
User is offline View Users Profile Print this message

Author Icon
DLJNA
Apprentice (20-49)

Posts: 29
Joined: 11/17/2005

Thank you so much for jumping in Hamid (everyone's welcome!).

This is precisely the kind of info I need. We will try this approach, so that only these servers (DNS clients) have a short entry TTL.

I really like this method.

Thank you so very much!

 Category Survey
 07/05/2013 11:29 AM
User is offline View Users Profile Print this message

Author Icon
Ricardo Castaneda
HelpSystems

Posts: 3424
Joined: 01/11/2008

Hi DLJNA,

Thank you for the update.

 Category Survey
AutoMate 8 version:
Windows version:


-------------------------

Ricardo Castaneda | Principal Support Analyst
AutoMate | Division of HelpSystems
HelpSystems, LLC | T: +1.213.738.1700
Statistics
18258 users are registered to the AutoMate Discussion forum.
There are currently 0 users logged in.
The most users ever online was 5551 on 01/08/2018 at 11:11 AM.
There are currently 752 guests browsing this forum, which makes a total of 752 users using this forum.

FuseTalk Enterprise Edition v4.0 - © 1999-2020 FuseTalk Inc. All rights reserved.

Sitemap Network Automation Software Blog