cancel
Showing results for 
Search instead for 
Did you mean: 
Create Post

Orion Platform 2018.4 Centralized Upgrades Walkthrough

Product Manager

Network Performance Monitor (NPM) 12.4 and the Orion Platform 2018.4 are now generally available in your customer portal. For those of you subscribing to the updates in What We're Working on for NPM (Updated June 1st, 2018)  you may have noticed a line item called "Centralized Upgrades." This update will give you the first chance to experience Centralized Upgrades on your environment.

Great news this upgrade is going to be easier than ever!
pastedImage_0.png

Planning for Your Upgrade to 2018.4

Read the release notes and minimum system requirements​ prior to installation as you may be required to migrate to new server or database infrastructure. For quick reference, I have provided a consolidated list of release notes below.

Note: Customers running Windows Server 2012, 2012 R2, and SQL 2012 will be unable to upgrade to these latest releases prior to migrating to a newer Windows operating system or SQL database version. Check for the recommended Microsoft upgrade path through the upgrade center.

See more information about why these infrastructures are deprecated here: Preparing Your Upgrade to Orion Platform 2018.4 and Beyond - Deprecation & Other Important Items

SolarWinds strongly recommends that you update to Windows Server 2016 or higher and SQL Server 2016 or higher at your earliest convenience. 

Refresh your upgrade knowledge with the following upgrade planning references.

Always back up your database and if possible take a snapshot of your Orion environment.

Start Your Upgrade on the Main Polling Engine

Download any one of the latest release installers to your main polling engine.

pastedImage_12.png

For the screenshots that follow I'm upgrading my Orion deployment with the following setup:

  • Main Polling Engine is installed with Virtualization Manager (VMAN) 8.3 and will be upgraded to VMAN 8.3.1
    • Utilizes a SQL 2016 database
  • Three scalability engines
    • One Free Additional Polling Engine for VMAN on Windows 2012
    • One Free Additional Polling Engine for VMAN on Windows 2016
    • One HA Backup on Windows 2016

My first screen confirms my upgrade path to go from 8.3 to 8.3.1.

  • If I'm out of maintenance for a specific product, I would see indicators here first on the screen. Being out of active maintenance will prevent you from upgrading this installation to the latest, so please pay attention to the messaging here.
  • The SolarWinds installer will upgrade all of the products on this server to the versions of product that are compatible with this version of the Orion Platform for optimal stability. This may mean that you'll be upgrading more than just one product.
  • When in doubt, feel free to run the installer to see the upgrade path provided, so you can plan for your downtime. Cancelling out at the pre-flight check stage will give you all the information needed to plan ahead, without surprises and without changes to your environment.  This information can also be used for your change request before scheduling downtime for your organization.

pastedImage_19.png

The second step will run pre-flight checks to see if anything would prevent my upgrade from being successful on the main polling engine.

  • In case there are no blocking, warning, or informational pre-flight checks, we will proceed straight to the next step, accepting the EULA.
    • My main polling engine server and DB meet all infrastructure system requirements for the 2018.4 Orion Platform, so I am not shown any blocking pre-flight checks at this stage.
  • Pre-flight checks can block you from moving forward with your installation
    • You  may need to confirm whether you meet new infrastructure requirements (e.g. NTA 4.2.3 -> 4.4 upgrade) to proceed. Blockers will prevent you from successfully installing or upgrading, so the installer will not allow you to proceed until those issues have been resolved. 
    • Warning pre-flight checks give you important information that could affect the functionality of your install after upgrade but will not prevent you from successfully installing or upgrading. 
    • Informational pre-flight checks give you helpful troubleshooting information for "what if" scenarios, in case we don't have enough information to determine whether this would be an active issue for your installation.

The online installer will start to download all installers needed from the internet

  • SolarWinds recommends that you use the online installer because it will be able to auto-update and download exactly what's needed for the installation. Not only is it more efficient, but it will save you from downloading unnecessary or outdated bits.

This screen gives you an overview of next steps. The Configuration wizard will launch next, to allow you to configure database settings and website settings.

In this release, all scalability engines, including Additional Polling Engines, Additional Websites and HA Backup Servers, can be upgraded in parallel manually, using the scalability engine installer. Manual upgrades are still supported, but if you have scalability engines, please try our centralized upgrade workflow to save you time.

pastedImage_0.png

Follow the configuration wizard steps to completion. If you only have a main polling engine to upgrade, your installation is now complete. Log in to your SolarWinds deployment and enjoy the new features that have been built with care for your use cases.

Centralized Upgrades of the Scalability Engines

For those customers who have chosen to scale out their environment using scalability engines, such as Additional Polling Engines, HA Backup Servers or Additional Websites this is the section for you.

If you kept the "Launch Orion Web Console" checkbox checked in the final step of the Configuration Wizard, the launched web browser session will navigate you directly to the Updates Available page, where you can continue with the Centralized Upgrade workflow. If you want to open a new web browser session on a different system, you can quickly navigate to where you want to go by following these steps.

Launch the web browser and log in.

pastedImage_3.png

Navigate to 'My Orion Deployment' from the Settings drop-down.

pastedImage_4.png

Click to the UPDATES AVAILABLE tab. If this tab is not showing, that means there are no updates available for you to deploy.

pastedImage_6.png

Click Start, to begin the process of connecting to your scalability engines.

pastedImage_7.png

My environment is not experiencing any issues connecting to my scalability engines.

pastedImage_8.png

Bookmark this page Connection problems during an Orion Deployment upgrade - SolarWinds Worldwide, LLC. Help and Support  for future guidance on common "gotcha" scenarios, and how to handle them.

After the contact with scalability engines has been established, pre-flight checks will be run against all scalability engines

pastedImage_9.png

Looking at my pre-flight checks you can see that one server PRODMGMT-49 has a blocker that would prevent upgrades from occurring, mainly that it does not meet infrastructure requirements for this version of the Orion Platform.

pastedImage_10.png

However, my "Start Upgrade" is enabled. This is because if at least one scalability engine is eligible for upgrade, we will allow you to proceed. Only when none of the scalability engines are eligible will this button be disabled. Pay attention to servers that have blocking pre-flight checks, as you will have to manually upgrade them or move items being monitored via this scalability engine to one that is upgraded.

Clicking "Start Upgrade" begins the centralized upgrade process, first by downloading all the necessary bits to all the scalability engines in parallel. Notice how my scalability engine that was on incompatible 2012 infrastructure is not being upgraded.

pastedImage_11.png

Grab a coffee as the rest of your installation and configuration happens silently on each of the servers being centrally upgraded.

pastedImage_12.png

pastedImage_13.png

Oh no, an error occurred. What can you do at this point?

  • Click Retry download after troubleshooting (e.g. did the scalability engine lose connectivity to the main polling engine?)
  • RDP directly into the server using the convenient RDP link that is provided

Common scenarios to investigate:

  • Is this scalability engine set up inconsistently from the other servers? For instance, you may have Engineer's Toolset on the Web installed on this server and not on the others.
  • Do some of the installed products have dependencies on .NET 3.5? Engineer's Toolset on the Web has a dependency on .NET 3.5 to be able to upgrade. Ensure that if you have enabled .NET 3.5 and try again.
  • Check the Customer Success Center for more scenarios to help while troubleshooting.

pastedImage_14.png

In my case, I clicked Retry and was able to get past the issue.

pastedImage_16.png

My upgrade is complete! Congratulations on an upgrade well done.

pastedImage_20.png

Click Finish to complete your Centralized Upgrade session.

Gotchas - What to do with Unreachable Servers

If your server isn't being blocked because of incompatible infrastructure, you have an opportunity to manually upgrade that server in parallel while the rest of your environment is being centrally upgraded.

In the installation example captured below, if I were to run the installer on the Additional Website that is currently being upgraded by Centralized Upgrades, I would be blocked from running the installer on that server. However for the listed unreachable Additional Website, I can run that upgrade manually with no problem in parallel.

pastedImage_26.png

If you're blocked from proceeding on a manual upgrade, you would see the following. Only until you have finished the Centralized Upgrade process will you be allowed to proceed with a manual upgrade that is blocked in this fashion. For these scenarios, simply navigate to My Orion Deployment and exit out of the deployment wizard flow to cancel the centralized upgrade session.

pastedImage_28.png

Manual Upgrades

Manual upgrades of your deployment are still supported. If you have only one scalability engine, Centralized Upgrades may not be the fastest way to upgrade. However, if you have more, it is. This upgrade is still beneficial for those considering using manual upgrades for their deployment, and the reason is the installation and configuration wizard process can now be run in parallel. Existing customers have always known that there were some scenarios where you could run the configuration wizard in parallel across servers (e.g. same server type) and some that you could not. It took time and training to understand what scenarios those were. In this release, that limitation is lifted, and all server types can be configured in parallel.

There are times where you may need to consider falling back to manual upgrades in combination with your Centralized Upgrade. As an example, take this installation: two have completed, one has the configuration wizard in process.

pastedImage_27.png

If the download, installation, or configuration is taking a long time for one of your scalability engines, and you need to see more information that is only available in the client, you may consider canceling out of the Centralized Upgrade session to resume the rest of your upgrade manually. The servers that have been upgraded thus far will remain in a good spot, so you can cancel out with confidence. Proceed with this option carefully, as you will want to ensure that you have upgraded everything by the end of your scheduled downtime.

pastedImage_32.png

Check the My Orion Deployment page to ensure that all the servers in your Orion deployment are upgraded.

pastedImage_31.png

Support

We have all been there, despite all the best intentions and all the preparation in the world, something went wrong. No worries! File a support ticket Submit a Ticket | SolarWinds Customer Portal  and start gathering diagnostics via our new web based and centralized diagnostics.

Click to the Diagnostics tab

Select all the servers in your deployment,

pastedImage_0.png

and click "Collect Diagnostics."

pastedImage_3.png

Sit back and relax as your diagnostics are centrally gathered in preparation for your support call.

Customer Experience

Early adopters and those who have participated in our release candidates have already begun to enjoy the benefits of centralized upgrades. Check out our THWACK forums for testimonials from customers just like you as they experience the new and improved "Easy Button" upgrade experience. Here's a link to one from one of our very own THWACK MVPs  The "Easy Button" has arrived with the December 2018 install of NAM (and other Sol... If you'd like to share your upgrades with me, I'm very interested, and we'd love to see screenshots and your feedback on this new way to upgrade your SolarWinds deployment.

More centralized upgrade success - Success with Centralized Upgrades

29 Comments
MVP
MVP

Would be great if I could get to that page but still nothing

Product Manager
Product Manager

dodo123  wrote:

Would be great if I could get to that page but still nothing

Does that mean when you navigate to the page, you're seeing a blank state? Do you have a support case open?

MVP
MVP

on the latest  version of npm

ie11

pastedImage_0.pngpastedImage_1.pngpastedImage_2.png

edge

pastedImage_3.png

chrome

pastedImage_4.png

ticket # 00203495 but was closed as wasnt getting anywhere with support saying having an ssl certificate is not supported.

i managed ti get it working under http before i last upgraded but not good enough and not secure

Product Manager
Product Manager

  wrote:

on the latest  version of npm

ie11

pastedImage_0.pngpastedImage_1.pngpastedImage_2.png

edge

pastedImage_3.png

chrome

pastedImage_4.png

ticket # 00203495 but was closed as wasnt getting anywhere with support saying having an ssl certificate is not supported.

i managed ti get it working under http before i last upgraded but not good enough and not secure

Thanks I'll take a look at the ticket.

Product Manager
Product Manager

serena  wrote:

dodo123   wrote:

on the latest  version of npm

ie11

pastedImage_0.pngpastedImage_1.pngpastedImage_2.png

edge

pastedImage_3.png

chrome

pastedImage_4.png

ticket # 00203495 but was closed as wasnt getting anywhere with support saying having an ssl certificate is not supported.

i managed ti get it working under http before i last upgraded but not good enough and not secure

Thanks I'll take a look at the ticket.

dodo123​ based on the ticket, it looks like support tried to contact you on Nov 22 to keep working on the issue, but received no response. Could you perhaps reopen the ticket, if support is not getting you the right solution, we can certainly reevaluate who is assigned to the case, but we want to make sure you're up and running to take advantage of this update.

Product Manager
Product Manager

serena  wrote:

serena   wrote:

dodo123    wrote:

on the latest  version of npm

ie11

pastedImage_0.pngpastedImage_1.pngpastedImage_2.png

edge

pastedImage_3.png

chrome

pastedImage_4.png

ticket # 00203495 but was closed as wasnt getting anywhere with support saying having an ssl certificate is not supported.

i managed ti get it working under http before i last upgraded but not good enough and not secure

Thanks I'll take a look at the ticket.

dodo123  based on the ticket, it looks like support tried to contact you on Nov 22 to keep working on the issue, but received no response. Could you perhaps reopen the ticket, if support is not getting you the right solution, we can certainly reevaluate who is assigned to the case, but we want to make sure you're up and running to take advantage of this update.

dodo123​ Try the steps here: The Deployment Health, Updates Available, and Diagnostics tabs are blank and diagnostics cannot be g...  and let me know if that helps your scenario please.

Level 10

Case # 00229309 - When trying to install VMAN additional collector/poller get message that unable to contact primary Orion Server

So what happened was that I was thinking of upgrading to the latest - I like to run the installer a few days beforehand so that I know it has made it through the pre-upgrade checks ok.

However, the installer knows it is an out-of-date installer and AUTO-updates it.  That's fine.  But right now I'm trying to install a VMAN Orion additional poller (replacement for VMAN appliance collector) and it would not work.

In the ticket is extensive information about what I tried and found.  The metadata for the installer, wherever that exists, points to a file SolarWinds.Orion.Installer-2.0.0.1463.exe, which is no longer in C:\ProgramData\SolarWinds\Installers on the primary poller, instead that AUTO-update replaced it with SolarWinds.Orion.Installer-2.1.0.1928.exe.  I went to my backups and put back the SolarWinds.Orion.Installer-2.0.0.1463.exe file and now things are going again.

So AUTO-updating files is great, but it can have unforeseen consequences.   And when I say AUTO-update I mean it just does it, doesn't ask if it's ok.  Which is fine, but as you could see from the above there are issues if you try to deploy anything after the installer has updated, but before you ran the upgrade all the way through.  This installer is a real gem, but even gems have rough edges to begin with.

Product Manager
Product Manager

jonathanswift  wrote:

Case # 00229309 - When trying to install VMAN additional collector/poller get message that unable to contact primary Orion Server

So what happened was that I was thinking of upgrading to the latest - I like to run the installer a few days beforehand so that I know it has made it through the pre-upgrade checks ok.

However, the installer knows it is an out-of-date installer and AUTO-updates it.  That's fine.  But right now I'm trying to install a VMAN Orion additional poller (replacement for VMAN appliance collector) and it would not work.

In the ticket is extensive information about what I tried and found.  The metadata for the installer, wherever that exists, points to a file SolarWinds.Orion.Installer-2.0.0.1463.exe, which is no longer in C:\ProgramData\SolarWinds\Installers on the primary poller, instead that AUTO-update replaced it with SolarWinds.Orion.Installer-2.1.0.1928.exe.  I went to my backups and put back the SolarWinds.Orion.Installer-2.0.0.1463.exe file and now things are going again.

So AUTO-updating files is great, but it can have unforeseen consequences.   And when I say AUTO-update I mean it just does it, doesn't ask if it's ok.  Which is fine, but as you could see from the above there are issues if you try to deploy anything after the installer has updated, but before you ran the upgrade all the way through.  This installer is a real gem, but even gems have rough edges to begin with.

Thanks Jonathan, I've grabbed the information you've sent, and am discussing with our internal team to see how we can improve this. By sending us this note, you're ensuring you and other customers will be getting an improved experience next time, so thank you very much for taking the time to report this.

Level 10

Thanks.   Going back to the centralised upgrades, I have now upgraded to;

Orion Platform 2018.4, NCM 7.9, NPM 12.4, DPAIM 11.1.1, NTA 4.5.0, VMAN 8.3.1, SAM 6.7.1, NetPath 1.1.4

using centralised upgrades and everything appears to have gone through ok.   This is a very significant enhancement to deployment.

I have the primary poller and two additional pollers, with two collector pollers for Orion VMAN.  It did them all.

Level 10

Ah, spoke too spoon.   Following the upgrade Screen Personalisations don't work any more - I can't see widgets.

Also, on the Virtualisation Summary screen my custom query list has gone - wiped out by the upgrade.

I'd check things and let you know about some of the things you've got wrong and which don't work but the last beta wouldn't even install, and 7-8 times per year I have to set up a test system to investigate and check new features.

It would be better if Solarwinds operated some type of developer / test system, which had limited capabilty, but which could be upgraded.   When you re-create systems from released release candidates you forego testing of systems where you had made many customisations.   I mean, who knew that the upgrade would wipe out my customisations on this screen - I never expected it to.

Anyway, new case

Case # 00230010 - Unable to modify page using Page Customizations - unable to see widgets.

Level 10

NOTE.  Reporting also no longer works - you don't get the list of objects to select, i.e. create a report with a table.

I'd have to recommend steering clear of upgrades right now.  I was pushing mine before our change control freeze tomorrow, but wish I hadn't bothered now.  I can see I will be on the phone to Solarwinds for the rest of the week trying to figure out what's gone wrong.

Level 10

Ok, so tried with Chrome and those features are working.   It's a bit ridiculous since I logged 00119529 which I later closed because support didn't appear interested.  In that older ticket I had these same problems with Chrome, so I switched to using IE, Firefox and Edge.  Now those browsers don't appear to work but Chrome does.   So, .... back to Chrome then.

There was a time when this product works fine with all the browsers.  From now on perhaps it would be wise to let me know in the release notes which versions of which browsers the produce has been tested with.

Product Manager
Product Manager

jonathanswift  wrote:

Ok, so tried with Chrome and those features are working.   It's a bit ridiculous since I logged 00119529 which I later closed because support didn't appear interested.  In that older ticket I had these same problems with Chrome, so I switched to using IE, Firefox and Edge.  Now those browsers don't appear to work but Chrome does.   So, .... back to Chrome then.

There was a time when this product works fine with all the browsers.  From now on perhaps it would be wise to let me know in the release notes which versions of which browsers the produce has been tested with.

We still support those browsers. I'm following the case that you've created, 00230010 to see if we can get a look at your diagnostics. We have seen quite a few upgrades and this is something new, so I'm interested in seeing what the root cause is. Thanks for your continued transparency into your upgrade.

Product Manager
Product Manager

serena  wrote:

jonathanswift   wrote:

Ok, so tried with Chrome and those features are working.   It's a bit ridiculous since I logged 00119529 which I later closed because support didn't appear interested.  In that older ticket I had these same problems with Chrome, so I switched to using IE, Firefox and Edge.  Now those browsers don't appear to work but Chrome does.   So, .... back to Chrome then.

There was a time when this product works fine with all the browsers.  From now on perhaps it would be wise to let me know in the release notes which versions of which browsers the produce has been tested with.

We still support those browsers. I'm following the case that you've created, 00230010 to see if we can get a look at your diagnostics. We have seen quite a few upgrades and this is something new, so I'm interested in seeing what the root cause is. Thanks for your continued transparency into your upgrade.

jonathanswift​ have you perhaps refreshed your IE browser cache? One of the theories that we have internally while we were talking about your case is that your browser cache has references to something old and that's why it's working for you in Chrome but not IE. Please try it and let us know.

Level 10

Diagnostic files uploaded.   However, it does work in Chrome.  If you were a Chrome-only environment then no problem, but since users tend to use a variety of different browsers it's an issue.   Some environments won't even allow Chrome or Firefox but insist on users using IE as a corporate standard, to take advantage of Trusted Zones, etc.

IE = 11.125.16299.0

Update Version: 11.0.49 (KB4052978)

Chrome = 71.0.3578.98

Firefox = 64.0 (64-bit)

Edge = 41.16299.15.0, EdgeHTML 16.16299.

Level 10

Support fixed this, I can confirm.

1- Locate VOLUME:\ProgramData\SolarWinds\Orion\WebResourceCache.xml
2- Cut this file to the desktop.
3- Open CMD as administrator
4- Type IISRESET
5- Press Enter
6- Login to the web console and try to add resources again.

Now working on all browsers.   The upgrade took the main pollers/website down for about 10 minutes or so, and took care of all the other additional pollers.

Product Manager
Product Manager

jonathanswift  wrote:

Support fixed this, I can confirm.

1- Locate VOLUME:\ProgramData\SolarWinds\Orion\WebResourceCache.xml
2- Cut this file to the desktop.
3- Open CMD as administrator
4- Type IISRESET
5- Press Enter
6- Login to the web console and try to add resources again.

Now working on all browsers.   The upgrade took the main pollers/website down for about 10 minutes or so, and took care of all the other additional pollers.

Thank you, I've sent this information internally to our teams.

Product Manager
Product Manager

Did that get you going?

MVP
MVP

Smooth process - successful upgrade to 2018.4 with centralized upgrades. Thanks for the write-up on the process.

Level 10

While I did have issues originally with the centralized upgrades, i believe those were self inflicted.  

I applied the recent hot fixes via centralized upgrades and it was flawless.

Much appreciated solarwinds team... 

Level 8

This is something we have looked forward to for a long time, but we just can't seem to get it to work. It always returns a list of servers that it says are unreachable. The administration service is running on the servers it says are unreachable, the administration log on the unreachable servers shows that it checks what is installed, all the required ports are open in both directions, DNS name resolution is working, I can even start/stop services on the affected servers from the Orion web console but for some reason it still reports some servers as unreachable. I had a case with support, it closed due to inactivity.

Level 10

serena​ , We have a fresh 2018.4 platform with NPM,NCM installed on it, all product licenses were applied in the month of December 2018, Now we have installed 2 APE's but license manager doesn't show these APE's due to which not able to license them, any pointer to this issue

Product Manager
Product Manager

b3nw00d  wrote:

This is something we have looked forward to for a long time, but we just can't seem to get it to work. It always returns a list of servers that it says are unreachable. The administration service is running on the servers it says are unreachable, the administration log on the unreachable servers shows that it checks what is installed, all the required ports are open in both directions, DNS name resolution is working, I can even start/stop services on the affected servers from the Orion web console but for some reason it still reports some servers as unreachable. I had a case with support, it closed due to inactivity.

Ben, I would love to get more details and get you going on centralized upgrades. Do you have the support case # that you filed? When you run a query on your database for table OrionServers what version is listed for the Administration service? Are all of your servers in sync with the main polling engine / are there any differences?

Product Manager
Product Manager

dhinagar_j  wrote:

serena  , We have a fresh 2018.4 platform with NPM,NCM installed on it, all product licenses were applied in the month of December 2018, Now we have installed 2 APE's but license manager doesn't show these APE's due to which not able to license them, any pointer to this issue

It looks like you have a support case on this issue, and the current recommendation is that you may have to do a manual registration of the ape licenses.

Level 10

serena​ ,  the issue got resolved, after we ensure APE's VIP able to resolve the Main polling engine's host name ( updating in /etc/hosts)

Level 10

We just did the 12.4 upgrade.  We ran into an issue and later found fix from some members on the the Thwack forum.  It would be helpful if all your support reps are familiar with this error.  We had to run the installation manually as a workaround:

NPM 12.4 - Error during contacting servers: Another Installation Session is still Active, you have ...

Product Manager
Product Manager

wlouisharris  wrote:

We just did the 12.4 upgrade.  We ran into an issue and later found fix from some members on the the Thwack forum.  It would be helpful if all your support reps are familiar with this error.  We had to run the installation manually as a workaround:

NPM 12.4 - Error during contacting servers: Another Installation Session is still Active, you have ... 

Thank you! Kacper is on our team and we've taken steps to update our support training documentation for our support reps to let them know of this error. I'm glad you're past the issue!

Level 10

Thank you.  If they can create a KB article it could save support calls.  It's a very easy fix.

Level 10

We encountered an issue where Centralized Updates would throw an error downloading a specific file on the Polling Engines for the 12.5 HF2. This occurred on 7 out of 8 Polling Engines, with no apparent pattern. Copying that specific file over directly from the Main Poller to the Installers folder on the Polling Engines solved that issue. All other installer files were downloaded just fine.

pastedImage_0.png

We have used the Centralized Updates before to go from 12.5 RC to 12.5, with no issues at all.

About the Author
Serena has been a software developer, program manager, product owner, and has found her passion in being a product manager. She loves building solutions for customer problems. Currently she's a Senior Product Manager at SolarWinds, enthusiastically responsible for Server & Application Monitor.