Upgrading AMDGPU 18.50 to 19.30 on Ubuntu 18.04

Hope you find this post helpful. My system is running Ubuntu 18.04.3 LTS with an AMD Ryzen 5 2400g and a separate Radeon RX 570 GPU.

Steam gameplay isn't perfect, but it's certainly good enough for my needs (minor screen tearing on most games, but it's not to the point of being overwhelmingly annoying or distracting).

Running sudo apt list --installed |grep -i amdgpu-pro reveals that I'm running the amdgpu-pro driver version 18.50-708488

Amdgpu 19.30 was released November 5th 2019 and can be found here: https://www.amd.com/en/support/kb/release-notes/rn-rad-lin-19-30-unified and is only compatible with the following distros/versions:

• Ubuntu 18.04.3
• RedHat Enterprise Linux 8.0
• RedHat Enterprise Linux 7.6
• CentOS 8.0
• CentOS 7.6
• SLED/SLES 15 SP1

Performing the Upgrade

1. Backup Backup Backup. Be sure you have proper backups and restore mechanisms in place. For my system, I use multiple backup solutions including Timeshift, as well as daily, weekly, and monthly Rsync backups. 

2. Download the 19.30 driver for your system (Ubuntu 18.04.3 version available here)

3. Extract the archive, either using Ubuntu's GUI or using tar -Jxvf

4. CD into the resulting directory and run the installer:  sudo ./amdgpu-pro-install

You'll find that the installer will list a number of packages to be upgraded. Go ahead and choose the default Y by hitting Enter and it will progress through the upgrade. Note you'll see it "Deleting module verison 18.50" during the 19.30 installation.

Once complete, you won't get any obvious "Complete!" or "Success!" prompts, but will rather be taken back to the command prompt. If you scroll up a bit, you should see the message "DKMS: install completed."

Assuming you don't see any errors mentioned, go ahead and reboot. You can verify 19.30 is the current driver by running sudo apt list --installed |grep -i amdgpu-pro again. 

Vulkan

Now, keep in mind that in order to use the Vulkan driver, Vulkan must be SDK needs to be at version 1.1.109.0, per the official documentation above. Running vulkaninfo should show you what Vulkan Instance Version you have. In my case, I had 1.1.70 installed, so an update was needed. You'll find that 1.1.109 isn't available over at https://vulkan.lunarg.com/sdk/home. For the meantime, I have simply installed the latest SDK at version 1.1.126 (the latest available at the time of this writing). If you run into this issue as well, please upvote and comment over here: https://community.amd.com/message/2941548

Good luck!

Read More

The vRealize Operations SSL Certificate Expiration Issue

[UPDATE #2 Nov. 18 2019] The KB is available again. Please subscribe to the KB for updates regarding the available .pak files, and engage VMware GSS if you have further questions/concerns.

[UPDATE #1 Nov. 18 2019] The linked KB article below has been pulled from public consumption due to an issue found in the mentioned .pak file. I will update this post as soon as the KB is re-published, or a new article is posted.

The VMware Product team has been notifying customers about an issue that could potentially impact your vROPs instance, causing the tool to become inoperable. See the notice and summary below. The key take away here is the initial install date. Please reference VMware KB 71018 for more information and steps to remediate.

The notice states, "Customers that have installed vRealize Operations v6.x and above may be impacted by the internal SSL Gemfire certificate expiration on November 19, 2019. The internal certificates for vRealize Operations Manager are set to expire five years after its initial installation. For example, if vRealize Operations 6.0.0 was installed on November 19, 2014, and incrementally upgraded to later versions, the internal SSL certificate will expire on November 19, 2019.

There is a required PAK file update that has been released for customers running vRealize Operations v6.x and above that renews the certificate to prevent it from going into system inoperability."

To summarize:

• All vROPs instance 6.0 and newer are impacted
• There is no fix for versions 6.0 - 6.2. You must upgrade to 6.3 or higher.
• The internal Gemfire SSL cert expires 5 years after the initial install of vROPs. The first of these will start happening November 19th. 
• When the cert expires, you cannot view any data, though collections will continue to run.
• Upgrades to vROPs have no impact on the cert, so the initial expiration date will still apply, even if you've upgraded vROPs several times since installation.

Read More

VMworld 2019 US Day 2

VMworld US 2019 Day 2 has already come and gone, and with it, more developments and exciting action. 

Day 2 Keynote Speakers  

Ray O’Ferrell – EVP, Emerging Cloud Native Apps BU, VMware
June Yang – VP, Product Management and Engineering, VMware
Macros Hernandez, Chief Technologist – Networking and Security, VMware
Purnima Padmanabhan, VP of Products – Cloud Management, VMware
Shikha Mittal, Director, Product Management & Strategy, VMware
Greg Lavender, SVP & CTO of Cloud Architecture, VMware
Sanjay Poonen – COO, VMware
Lindsey Vonn, Olympic Gold Medalist
Steve Young, Record setting professional NFL Quarterback

Day 2 Keynote Summary 

Today’s General Session was hosted by Ray O’Ferrell, VMware’s recently former CTO and now EVP leading VMware’s expanding Cloud-Native Apps BU. The session started with a video of a young college girl in the not too distant future leveraging dozens of innovative applications on her phone that are critical to her studies, travel, and hobbies all on top of on a robust, secure cloud platform built on VMware. 

Ray started the keynote discussion on our world of digital transformation and digital infrastructure, stating the fact that every single industry is touched by this digital transformation. The enterprises that leverage all this new technology are quickly and pragmatically going to be the enterprises that can compete in the digital future. 
He then introduced a fake online company for the purposes of this session, Tanzu Tees, to help the audience understand how VMware can help in each of the four categories above. How does Tanzu Tees manage multiple frameworks, clouds, and multiple languages to deliver the necessary experience to their customers? Ray broke it down into four categories:

1. Build and Run 
2. Connect and Protect
3. Manage
4. Experience

Build and Run – Ray talked about building cloud native applications and microservices utilizing VMware services like Tanzu, Pivotal, and Bitnami. He then introduced June Yang onto the stage to give a demo on the ‘run’ part of ‘build and run’. She showed off a demo of the former ‘Project Dimension’, now called VMware Cloud on Dell EMC. She ordered a new SDDC location for Tanzu Tees. When the order is placed, VMware and Dell EMC start building a fully configured and managed SDDC rack for the customer at the location specified. She then demoed Project Pacific, which Pat introduced in the keynote yesterday. Finally, she demoed Tanzu Mission Control, which showed how we can manage multiple Kubernetes clusters in multiple clouds all from one location.

Connect and Protect – For this next demo, Ray introduced Marcos Hernandez. Marcos showed how we can configure service level objectives for our applications in Tanzu Mission Control, and how AVI Networks load balancers can reroute traffic to reduce downtime. He then showed how NSX Service Mesh can auto scale workloads to bring our applications back into SLO compliance. In his second demo, Marcos introduced NSX Intelligence. NSX Intelligence was able to drill down into our applications and apply security recommendations based on multiple data inputs. He showed how NSX Intelligence can simulate a security policy before applying it, avoiding unforeseen issues. Finally, Marcos demoed AppDefense and how it can be used to detect vulnerabilities in your SDDC and drill down to the offending application or process to remediate quickly.

Manage – For the manage portion of the keynote, Ray introduced Purnima Padmanabhan to demo consistent management with VMware. She started by showing how Wavefront can help us pinpoint problems through its event collection and analysis capabilities. Wavefront will pull information from NSX Service Mesh to map dependencies and show application performance. She seamlessly switched from Wavefront to vRealize Operations to troubleshoot a VSAN issue. In vRealize Operations she enabled Magna for the vSAN cluster, which allowed vRealize Operations to automatically tune the cluster for read performance without the administrator having to deal with the dials and knobs. In her next demo, Purnima demoed how CloudHealth can help track cost for our applications and give recommendations on how to lower those costs.

Experience - Shikha Mittal came onstage for the Experience portion of the keynote. She gave a demo of VMware’s digital workspace solutions for quick employee onboarding. She announced VMware Horizon Desktop as a service on AWS and demoed how to quickly deploy a new Horizon site in AWS. She then showed off the Horizon Image Management Service, which allows admins to view a global catalog of images to deploy. Administrators can use policies to route different users to different sites and customize the applications that they have access to. She then showed how Workspace One can manage things like app issues, device issues, and overall user experience on all connected devices. Finally, she demoed how Workspace One Risk Analytics uses machine learning to learn expected behavior and protect high risk users from malicious software. 

“Art of the possible in a multi-cloud world” 
For our final demo, Ray introduced Greg Lavender to discuss customer concerns with their journey to the cloud and how to refactor or migrate their current applications. Greg talked about how Bitfusion’s technology lets us leverage GPUs and FPGAs to run machine learning and NFV tasks in the cloud or on-prem. He then moved onto a cool demo where he showed how we can use VMware HCX to migrate an IoT edge application from Azure’s IoT platform to AWS’s IoT platform and vice versa, with zero network downtime thanks to NSX. 

After the demo, Pat Gelsinger came up on stage and announced that Ray O’Farrell would be moving on from his CTO role to leading VMware’s cloud native division, and that Greg Lavender would be taking Ray’s old spot as CTO. You can read more about this surprise announcement here. 

General Session Day 2 Replay

Summary of Day 2 Announcements

The Next Chapter in Cloud Innovation: VMworld Day One Recap
The first day of VMworld was exciting. Read VMware Vice President and CTO Chris Wolf’s recap of the day in his blog, VMworld 2019 and Our Relentless Pursuit of the Possible. Innovations and enhancements in hybrid cloud, VMware’s Tanzu Portfolio, VMware Cloud on AWS, and much more demonstrate our commitment to creating value for our customers.  

Watch VMworld General Sessions On-Demand
Monday’s General Session with VMware CEO Pat Gelsinger and COO Sanjay Poonen is now available on-demand. Watch the replay and explore bold innovations that the VMware ecosystem is bringing to market, with a special focus on new breakthrough offerings at the intersection of cloud, apps, containers, networking, mobility and security. Plus, learn how you can hone new skills and harness new game-changing technologies.

On Tuesday, VMware CTO Ray O’Farrell guided a conversation with a series of hands-on demos that look at the dynamic interplay between apps, infrastructure, cloud, networking, security, and devices. Then, Sanjay Poonen hosted two professional athletes, skier Lindsey Vonn, the top US skier of all time, and Steve Young, Pro Football Hall of Famer and former 49er quarterback. They share personal stories of their unrelenting work ethic, how they overcame obstacles, and what drives their passion.  

The Industry’s Only Complete Software-Defined Networking and Security Stack Built for the Multi-Cloud Era
Today we also announced a broad set of networking and security innovations and major milestones that further advance our vision of the Virtual Cloud Network. Important updates include the introduction of VMware NSX Advanced Load Balancer; new and enhanced network and security analytics capabilities, delivered through VMware vRealize Network Insight 5.0 and NSX Intelligence, which build upon advancements of VMware’s SD-WAN by VeloCloud, the industry’s only hyperscale cloud networking solution. Read the announcement to learn more. 

Introducing Project Magna
We also provided a preview of Project Magna, which is a reinforcement learning-based automated vSAN performance tuning service. Project Magna will be the first instantiation of the self-driving data center vision, beginning with VMware vSAN. It will focus on using reinforcement learning to understand application needs and to continually and dynamically auto-optimize the underlying vSAN configuration to drive greater performance and efficiencies. Magna will be delivered as a combination of a VMware cloud service integrated with vRealize Operations.

Expanding Telco and Edge Cloud Portfolio to Enable Better Connectivity and Automation for Communication Service Providers and Enterprises 
VMware has expanded its Telco and Edge Cloud portfolio to drive real-time intelligence for telco networks, as well as improved automation and security for telco, Edge and IoT applications. This is an important development and one that enables them to efficiently deploy, realize and monetize their 4G and 5G network investments through an expanded set of use cases targeting enterprise customers. We also announced the closing of the acquisition of Uhana. Read the press release for more information.

Read More

VMworld 2019 US Day 1

VMworld US 2019 Day 1 is in the books along with a slew of exciting announcements and developments. Expanded partnerships in hybrid cloud, VMware Tanzu preview, and enhancements to Workspace ONE are just a few of the major announcements that help position VMware as a critical player in today's Enterprise.

Day 1 Keynote Speakers (in order)

Pat Gelsinger, CEO, VMware
Joe Beda, Principal Engineer, VMware
Jensen Huang, CEO and Founder, NVIDIA (video segment)
Hans Vestberg, CEO, Verizon (video segment)
Sanjay Poonen, COO, VMware
Rathi Murthy, COO, Customer Operations, Gap Inc.
Timothy Snyder, Deputy CTO, Freddie Mac

Day 1 Keynote Summary

Today at the VMworld 2019 Day 1 General Session, we showcased our vision of a software architecture that enables any app, on any cloud, to any device with new technology offerings, customer stories and strategic partnerships. Pat Gelsinger provided more detail on our strategy to accomplish this goal with the announcement of the VMware Tanzu Portfolio, as well as innovations such as Project Pacific, and our expanded Cloud Offerings and collaborations such as VMware Cloud on AWS and the recently announced NVIDIA partnership. Following Pat, Sanjay Poonen interviewed leading-edge customers, including a video interview with several executives at Verizon and onstage interviews with Gap and Freddie Mac executives; all of whom are harnessing game-changing technologies to drive innovation across their companies and industries. Sanjay also unveiled innovations to our Workspace ONE platform that are enhancing the employee experience from hire to retire with industry-first concierge services and advanced management and security features. 

General Session Day 1 Replay

Summary of Day 1 Announcements

VMware Delivers a Hybrid Cloud Platform Powering Next-Generation Hybrid IT

VMware is continuing to build its agnostic platform while expanding partnerships. No matter the platform you choose now or in the future (AWS, Dell, Microsoft, Google, IBM, and HPE), VMware continues to be strongly positioned to be the provider for any cloud, as well as on-prem—we use the same software and VMware Validated Design with our VMware Cloud Foundation (VCF) stack, enabling you to migrate, modernize and manage workloads across clouds, data centers, and edge locations with consistent hybrid cloud infrastructure and operations. 

We’ve also made a lot of complementary announcements on automation, disaster recovery as-a-service, and proactive support, among others. Read the announcement for additional information. 

VMware Cloud on AWS Helps You Migrate and Modernize Applications
A new study shows that customers who want to migrate to the cloud can save 59 percent in operational costs with VMware Cloud on AWS versus the equivalent capacity in a traditional data center. Additionally, VMware Cloud on AWS is now available in 16 AWS regions globally and can be purchased from both AWS and VMware. Check out the press release for the full study and additional resources.  

NVIDIA and VMware Intend to Deliver Accelerated GPU Services for VMware Cloud on AWS The partnership is designed to power modern enterprise applications, including Artificial Intelligence (AI), Machine Learning (ML) and data analytics workflows. These services will further allow you to migrate workloads that previously were not well-suited for the cloud and take advantage of high-performance computing, machine learning, data analytics and video-processing applications. 

Simply put, we’re adding capabilities that will make it easier to move specific and graphic-based workloads to the cloud. Read the announcement for the details.

VMware Announces VMware Tanzu Portfolio 
While this is a preview into what’s coming, Tanzu enables you to build modern apps on Kubernetes and manage all of your clusters from a single control point. Ultimately, integrated teams can request what they need, when they need it, and get it effectively and efficiently. This helps you keep up with the speed of tech while not having to worry about changes and patching continuously. 

Additionally, take a look at Project Pacific, which will enable IT operators and developers to accelerate innovation by running modern applications on VMware vSphere with native Kubernetes. Read the release for details. 

VMware Unveils Innovations Across Its Industry-Leading Workspace ONE Platform 
New enhancements and capabilities to Workspace ONE will help you grow, expand and transform. The new modern management, security and multi-cloud VDI innovations will allow IT teams to leverage the power of automation to intelligently manage and secure access to any app, on any cloud, delivered to any device. And it makes you faster. You have the tools you need, and they’re integrated across multiple devices and platforms, such as Dell.

For example, VMware is leading the way in helping IT and HR reimagine the onboarding and day one employee experiences with the introduction of digital concierge services, leveraging the extensive automation capabilities of the Workspace ONE platform and the richness of partner integrations. Check out the article for more information.

Finally, you may want to read through VMware CTO Chris Wolf’s summary of the day: VMworld 2019 and Our Relentless Pursuit of the Possible. This provides you with a consolidated look at the vision and strategy behind today’s announcements. 

Read More

Upgrading Seafile Server from version 6 to version 7

In this post I'll cover the process of upgrading Seafile Server from 6.3.4 to 7.0.4. I've been running Seafile Server in an Ubuntu VM on my Linux Server for about a year and a half. Along with using Let's Encrypt's Certbot, it's been an incredibly stable and virtually 'hands off' service. 

Pre Upgrade

1. Backup current system. As I run Seafile in a VMware Workstation (for Linux) VM, it's easy to create a quick snapshot of the current system and configuration that I can revert to in the case of any unforeseen issues. Additionally, I run Timeshift with daily/weekly/monthly rollups of my entire server. Can't ever be too careful!
2. Download the new Seafile bits. Seafile Server is a quick download from their download page, and is simple to run a wget on the download to get it to your server. 
3. Make sure the current server is functional. Before you perform any upgrade of any service/application, it's best practice to ensure the current iteration is up and running and functional. This is a practice any IT professional should follow regardless of your employer/application/service in question. Failing to do so can bring unwanted behaviors to your updated instance, or even cause the upgrade to fail altogether.
4. Review the Release Notes. Once again, a best practice anyone in IT should follow before jumping into an upgrade is reviewing the Release Notes of the target version. Release Notes are known to include critical information not only related to new features, but often around the upgrade process itself.

Upgrade

1. After reviewing the release notes and downloading the 7.0.4 bits to your server, untar the file using the command tar -xzvf seafile-server_7.0.4_x86-64.tar.gz and ensure the resulting folder is in the same directory as your current seafile installation
   
2. The next step is to stop your seafile services (seahub and seafile) by running ./seahub.sh stop && seafile.sh stop
3. Once the services are stopped, go into the new 7.0.4 directory and find the upgrade scripts under the upgrade folder. Starting with the script of your current server (in my case, 6.3.4 would correlate to upgrade_6.3_7.0.sh), run each script
   
   
   
4. Run the upgrade script and press ENTER
   
   
   
5. You may find some [info] and [warning] messages, but as long as there are no [error] messages, you'll look for "Upgraded your seafile server successfully"
   
   
   
6. Ensure your seafile and seahub services start successfully, then verify your version on the server in the About section. 
   
   
   
   
   
   
   
   
   And congrats you're done! 
   
   
   One thing I tried testing that doesn't seem to be fully working yet is editing the new Published Library's .md file. I get a spinning circle as if it's thinking or loading something. In my case, it just spins indefinitely:
   
   At any rate, your server should now be upgraded! Let me know how your upgrade went in the comments. Thanks for reading!
   
   

Read More

Troubleshooting the Zmodo Greet Pro Smart Doorbell

Earlier this year I bought a Zmodo Greet Pro Smart Doorbell. It was cheap, didn’t require a monthly subscription, and had relatively decent reviews. It was a breeze to install, and had it up and running the day it arrived. Fast forward about 1.5 months, and something weird happened.

I was on vacation, enjoying Yellowstone in the Spring, and noticed while I was out that my home network wasn’t working. I couldn’t connect to Plex, couldn’t connect to my OpenVPN server, etc. My roommate notified me soon after and had to reboot the router to get everything working again. An annoying issue, but a simple fix.

Upon returning back home, I noticed everything working as expected, except my Doorbell was flashing a green ring. I looked this up and the Zmodo documentation says that indicates it’s ready to be setup. So I checked the Zmodo app and sure enough, it showed my doorbell, which was previously configured and working fine, as “Device Disconnected.”

So I removed the device from the app, and attempted to set it up again from scratch. However this is where the frustration set in. Every attempt to add it back via QR code timed out and didn’t work. I thought cutting power to the entry way could help reset the thing, but that didn’t do anything. And to my surprise, it was still powered on when I killed the power. I just learned this thing has a battery.

Next I remove the thing off the wall, remove the back panel in an attempt to find the battery and disconnect it. Turns out I couldn’t locate that, so I just left it and waited for it to die on it’s own. It took about 15 or 20 minutes, but it eventually died out. So I put everything back together, plugged it back in on the wall and it now showed a solid blue ring. Documentation says this indicates it’s connected to WiFi. Sure enough, I see the device on my router. So I tried adding by QR code again to my app, and it still didn’t work!

This is when I found this Zmodo support article that mentioned how to add the device back to your app if it’s already connected to WiFi. You must add it by using the “Wireless Device and Smart Kit” option instead of the QR code option. Choosing “Wireless Device” from here, the app was able to locate the doorbell on my network and get it back into the app. Voila! I can now spy on the unsuspecting outside world.. that is on my front porch.

To summarize:
- Wifi going wonky can cause Zmodo Greet Pro to lose its mind, and flash a green ring
- When that happens, remove the doorbell and let its battery die all the way down (it will stop flashing green)
- After the battery dies, reinstall it and connect it to power. I’m guessing this would be all you need to do to get it working again. In my case, I had removed the device from the app before killing the battery
- If you removed it from the app like me, re-add it once the ring is a solid blue, by using the Wireless Device and Smart Kit option

Read More

vROPs for Horizon - Regain Compatibility and Insight into your VDI

If your business happens to utilize VMware Horizon for VDI, and vRealize Operations (vROPs) for infrastructure monitoring, insight, and capacity planning, you may also be a business that uses the vROPs for Horizon Adapter (V4H) to bring that level of insight into your virtual desktop infrastructure. That being the case, you likely noticed that good stretch of time where vROPs 6.7 broke compatibility with the existing vROPs Horizon Adapter (V4H) 6.5.1 forcing you to delay upgrades, or lose your vROPs insights into the Horizon environment until it was fixed.

So what's needed to bring everything to compliance and compatibility to get insight back into Horizon with vROPs? Ultimately, you need the 6.6 adapter and a patched vROPs instance. The V4H Adapter 6.6 now supports vROPs 6.7 and 7.0. And the process for patching and licensing the vROPs instance is outlined below.

Step 1:  Upgrade your V4H license

Head to the VMware Portal and find your vRealize Operations 6 Manager for Horizon license. Go through the process via the portal to upgrade it to vRealize Operations Manager for Horizon

Reference: How to upgrade license keys in My VMware

Step 2:  Patch vROPs

Follow the steps outlined in KB 60301 to apply the appropriate patch to your vROPs instance

NOTE: vRealize Operations Manager 6.6.1 GA, 6.7 GA, and 7.0 GA, do not include the Security Patch.

How can you tell if you already have this patch? 
If you're on version 7.0 and you see the patch below, you're still on GA and need to apply the security fix. Once it's been applied, you should see  build 7.0.0.11287812. Verify your working build per the KB linked above.

Step 3:  Install the new license 

Implementing the new key in your vROPs manager will bring compliance and compatibility once again. Here's the official documentation for applying licenses.

For the latest on vROPs and Horizon compatibility, always check the Interop Guide, and take a peek at KB 59651 as well.

Happy patching!

Read More