AnsweredAssumed Answered

Update pushes through SMACC working intermittently

Question asked by Samuel on Apr 3, 2019
Latest reply on May 8, 2019 by Gert-JanKosterman

Like • Show 2 Likes2
Comment • 17

Dear specialists,

I am having trouble regularly when pushing updates (agent or files) through the Inventory SMACC.

Often, the push never starts and I have to recreate it, several times in a row in some cases, before it eventually starts.

Has anyone faced this issue before?

This is annoying considering that I have to re-create those pushes every few weeks because the target collections do not update dynamically according to Gert/Martin

P.S.: I have asked Snow Support over a week ago but got no answer yet (CS0037401)

sean.moldovan

Apr 4, 2019 7:48 PM

Correct Answer

I have a question with regards to the filter that you are applying to the SMACC Agent Updates job: Is the filter applied to the job using a cloned View from an existing View?

One thing that can cause sometimes cause problems with the functionality of SMACC Agent Updates is when you apply a filter of a cloned of View. If you are using a cloned view for the filter target, please try creating a new view to apply to the filter to a Agent Update job. You can also try changing the method of filtering to something other than a view.

See the reply in context

1 person had this question

Outcomes

17 Replies

ester.memoli Apr 3, 2019 8:36 AM
Mark Correct
Correct Answer
Good afternoon Samuel,
thanks for raising your question with us.
I will check with Support what's the status of the mentioned case and how we can proceed with it.

Cheers,

Ester Memoli
Customer Success Manager EMEA
Snow Software srl - Milano - Italia
Welcome to Snow! Discover more here: Welcome to Snow – Where to start and things to know
5 people found this helpful
Like • Show 5 Likes5
Actions
Gert-JanKosterman Apr 4, 2019 2:48 AM
Mark Correct
Correct Answer
Hi Samual,

I'm sorry but I need to inform you ticket CS0021921 is closed this week. I again was asked for a remote session by Snow support since they were not able to reproduce the described scenario themselves. But since I have a workaround available (re-create rule) and did spend 'enough' time on this already I informed them they can close the ticket. But feel free to open one yourself and refer to CS0021921. I personally do not understand why they are not able to reproduce it themselves since it seems I'm not the only one running into this

Success,
Gert-Jan.
2 people found this helpful
Like • Show 3 Likes3
Actions
judith.rothwell Apr 4, 2019 3:28 AM
Mark Correct
Correct Answer
Hi Samuel and Gert-Jan
Thank you both for your comments. We are committed to finding the right solution to this issue and appreciate the time you dedicate working with us to achieve this.
I have raised the profile of CS0037401 and will work on this case with a view to finding the solution to help you both.
Thanks again for your feedback and Samuel my colleague will be in contact shortly.

Many thanks
Jude Rothwell
Customer Support Manager EMEA
Snow Software
3 people found this helpful
Like • Show 7 Likes7
Actions
- Samuel @ judith.rothwell on Apr 4, 2019 6:05 AM
  Mark Correct
  Correct Answer
  That was very quick!
  I was contacted this morning already. Because the error is difficult to reproduce, we could not sort it out today but if anyone faces that isse, what you can do is to set the log level to verbose on a client not receiving (pulling) the update. The error will likely be logged on the client side rather than the server side.
  Also, before that, check that the polling is set to reasonable values, obviously (defauld are interval=1200, variance=300)
  I will update this ticket and post as soon as I get some news.
  Thanks again
  4 people found this helpful
  Like • Show 5 Likes5
  Actions
sean.moldovan Apr 4, 2019 7:48 PM
Unmark Correct
Correct Answer
I have a question with regards to the filter that you are applying to the SMACC Agent Updates job: Is the filter applied to the job using a cloned View from an existing View?
One thing that can cause sometimes cause problems with the functionality of SMACC Agent Updates is when you apply a filter of a cloned of View. If you are using a cloned view for the filter target, please try creating a new view to apply to the filter to a Agent Update job. You can also try changing the method of filtering to something other than a view.
3 people found this helpful
Like • Show 5 Likes5
Actions
- Gert-JanKosterman @ sean.moldovan on Apr 5, 2019 2:01 AM
  Mark Correct
  Correct Answer
  We indeed had this issue with using a cloned view. But sometimes that's what you want. So the workaround is Ok but it should also work on cloned views. Sometimes a colleague created a view you later want to use for an agent update and then you have no clue if it's a cloned view or not?
  3 people found this helpful
  Like • Show 2 Likes2
  Actions
- Samuel @ sean.moldovan on Apr 5, 2019 2:22 AM
  Mark Correct
  Correct Answer
  Thanks Sean,
  I agree with Gert, I too use views I cloned.
  If this is a known issue, will a fix be made available?
  2 people found this helpful
  Like • Show 2 Likes2
  Actions
xwin Apr 5, 2019 11:03 PM
Mark Correct
Correct Answer
Agent update target devices based on assignedId, not by device name.
When we create Agent Update task, it will query and filter from device data and get the device's assignedId. The filter is based on the criteria that had been set.
If you faced with issues like when there is no response from the agent, you could start with this basic verification. Does the list of assignedId in your 'Agent Update Summary' is the same as assignedId currently in your 'All Device List'.

In my case, since some of my device keep on being updated/created with new clientId and assignedId.
So Agent Update will normally target previous inventoried devices and ignore the newly inventoried device.
3 people found this helpful
Like • Show 2 Likes2
Actions
- Samuel @ xwin on Apr 8, 2019 3:02 AM
  Mark Correct
  Correct Answer
  Hi Shahrin,
  This is a good advice thanks. I'm not sure exactly where you can see the "list of assignedId in your 'Agent Update Summary'". DO you have a screenshot?
  Also, I was originally talking of updates pushed to over 25,000 computers so although some might be in the situation you mention, the probabilities are high that the push should have worked on a high number of computers
  2 people found this helpful
  Like • Show 2 Likes2
  Actions
  - xwin @ Samuel on Apr 8, 2019 3:58 AM
    Mark Correct
    Correct Answer
    Hi Samuel,
    
    I have not worked with that kind of huge environment. I only work with around 8,000 devices at a time. Normally for intermittent issues, i would focus on server or db performance. Those two factors might contribute to the stability issue in high volume environment.
    I've been using default configuration for agent interval & variance.
    
    1) AssignedId verification.
    Below is the screenshots
    
    Agent Update Summary
    
    All devices list
    
    2) Creating Agent Update Failed
    I'm not really clear on what did you mean when you mentioned that 'push did not start'.
    In order to verify if the job task is successfully created, you may check your 'System Events'.
    Sometime, 'Filter job update' failed. End up 0 devices added to the update job.
    In this case, i actually have intermittent issues with my database
    
    3) Other possibilities
    Some other small issues that I've faced with agent updates.
    a) Server time is not up to date. - I set the update to previous dates. e.g. 1/1/1900. So it will start update immediately after the agent response.
    b) Wrong criteria set in Agent Update task. - I've set to the wrong site or agent type, resulting for 0 device added to Agent Update task. It may looks like the Agent Update did not run.
    c) High volume of agent connecting to Inventory Server. - I would restart inventory server services before doing an update. To allow more headroom for the agent to report back.
    d) Inventory Server not responding. - It happen previously to my environment, due to high volume and big sized snowpack being inventoried. Inventory Server actually hang and did not accept new agent connection. Thus no response for updates.
    The issue only solved after i cleaned up the devices data and update the agent config to prevent it to send big snowpacks.
    
    Hope this will give you some ideas.
    4 people found this helpful
    Like • Show 5 Likes5
    Actions
    - Samuel @ xwin on Apr 8, 2019 4:41 AM
      Mark Correct
      Correct Answer
      Thanks a lot for sharing that detailed description!
      That will be very helpful at th etime of troubleshooting updates, and probably not just for me!
      Great job
      2 people found this helpful
      Like • Show 2 Likes2
      Actions
    - sean.moldovan @ xwin on Apr 11, 2019 1:22 AM
      Mark Correct
      Correct Answer
      Excellent content. Great information. I am working on platforms of differing sizes - 100 node and 100k+, soon to be a 280,000 node Single Snow Software Platform (Enterprise Edition (EE) and Service Provider Edition (SPE)). Some additional things are often overlooked and you should watch out for:
      Snow Software System Requirements Met
      Make sure that the Microsoft SQL Server, Inventory Server and Snow
      License Manager meet the minimum software requirements. Under-resourcing SQL server, SLM and Inventory Server or having shared resources on a virtual infrastructure can degrade the performance.
      Keep up with the latest available version of Snow Software and Agents
      There have been some known issues uncovered and addressed with the release of new versions of software.
      Example: SIAW 5.0.2 had an issue which could cause the agent to crash and not restart making it impossible to use the SMACC for updating the agent as it was not polling in.
      For the example above, the SIAW needed to be uninstalled and reinstalled using a ITSM utility.
      Database Maintenance
      Another area that is often overlooked is the database maintenance. Most organizations have dedicated database administrators who take care of this functionality. Snow Software do not provide detailed instructions as to how maintain the SnowInventory and the SnowLicenseManager SQL databases. We defer to any SQL Server Maintenance to the freely available tools that are available from https://ola.hallengren.com/.
      Firewall and Snow Inventory Agent connectivity
      Matching TLS version settings for communications between Inventory Server - Master Server / Gateway Server and Snow Inventory Agent computer.
      Installing Snow Software Platforms on Virtual Computers
      Make sure that you do not overload any shared SQL servers and that you exceed the minimum Snow Software / MS SQL server system requirements.
      Virtual Hosts are regularly patched.
      Above is only a subset of what factors that can impact performance of when attempting to update Agent the Snow Inventory Agent using the SMACC.
      6 people found this helpful
      Like • Show 5 Likes5
      Actions
colin.campbell Apr 11, 2019 12:14 AM
Mark Correct
Correct Answer
"Working intermittently" is an accurate description!
2 people found this helpful
Like • Show 2 Likes2
Actions
Samuel Apr 26, 2019 10:23 AM
Mark Correct
Correct Answer
OK Today I finally recreated all my views from scratch instead of cloning.
All seems to be working perfectly now.

Known big bugs like that should be fixed please ... that bug has been around for very long.
2 people found this helpful
Like • Show 4 Likes4
Actions
- Gert-JanKosterman @ Samuel on Apr 26, 2019 1:53 PM
  Mark Correct
  Correct Answer
  Samuel, I completely agree on your big bugs statement. I created a ticket long time ago too but this bug is indeed still in...
  2 people found this helpful
  Like • Show 3 Likes3
  Actions
  - Samuel @ Gert-JanKosterman on May 8, 2019 9:23 AM
    Mark Correct
    Correct Answer
    Sad news: I made some more tests and found that our new views still exclude seeminly random endpoints.
    I re-opened a ticket to investigate and documented everything: CS0042877
    I realized our agent update push using the new view inexplicably excluded one server.
    I created a test view targetting this specific server and pushed the same update and this time the server is not excluded from the update.
    I will update the thread based on Snow Support's comments.
    3 people found this helpful
    Like • Show 2 Likes2
    Actions
    - Gert-JanKosterman @ Samuel on May 8, 2019 9:52 AM
      Mark Correct
      Correct Answer
      Hi Samual,
      
      I've got exactly the same issue. That's why each time I want to update something I completely create a new view + agent update rule. But this is very cumbersome and not the way you expect it to work...
      
      regards,
      Gert-Jan.
      2 people found this helpful
      Like • Show 2 Likes2
      Actions

Update pushes through SMACC working intermittently

Outcomes

Related Content

Recommended Content