Be A Citrix Hero During the Coronavirus COVID-19 Outbreak

I despise opportunity-based marketing. Yet, whether you consider the responses thus far to Coronavirus COVID-19 to be panic or sensible precautions, one thing is sure: There hasn’t been a demand for working from home worldwide like this since Bird Flu (or, was it Swine Flu?). Certainly not one this sudden and wide ranging. However, this case has been way worse in my observation; not medically but how our fear response has taken over. So how can you be the Citrix Hero during a season like this? If you have been listening to me on my podcast (Better Than You Found It) or reading posts on LinkedIn and such, you’ll be aware of a saying by which I live: You will not rise to the occasion. You will only rise to your level of preparedness. This holds true in EUC as much as in life.

You will not rise to the occasion. You will only rise to your level of preparedness. This holds true in EUC as much as in life. #CitrixHero Click To Tweet

Rising to the Coronavirus Occasion

If there is one thing I see consistently in Citrix Assessments – it is that the systems housing Citrix Virtual Apps and Desktop VMs (Single or Multi-Session OS) seem to be running past their maximums already. Adding on even more workloads suddenly is making this problem worse. Why? Because the IT systems were not PREPARED for this occasion. Time and again, I hear things like, “We’ll deal with that load when it comes.”

I hate saying I told you so.

How fast do you anticipate getting new servers given where the parts are manufactured? How will it get to you when delivery services start to become constrained? And if you believe the hype, how will you install them when you’re not allowed to go to work?

Citrix Hero Concept – Do More With Less

Something I talk about in my book Be A Citrix Hero is that with the right optimizations and design, you can do more with less. This is how YOU will be the Citrix Hero during the Coronavirus outbreak – you can set systems to perform at scale.

Tweet This to your friends and do everything you can to let your boss know: The same optimizations that improve user performance can help you trade performance for scalability during a crisis.

The same optimizations that improve user performance can help you trade performance for scalability during a crisis. #CitrixHero Click To Tweet

By organizing resources properly (Chapter 3), optimizing the operating system to eliminate waste (Chapter 1), Utilize RAM properly for PVS and MCS workloads (Chapter 4), and using entitlements such as Citrix Workspace Environment Management (Chapter 6), the Citrix Hero is able to do more with less and be prepared for temporarily exceeding the design specifications.

I tell a story in Chapter 6 of an environment that I had done all of these things, and during an administrative glitch, they managed to shut down eight of their twelve 7.15 LTSR Server OS VMs. We had recently implemented just the basic settings I describe for WEM. Imagine their shock that all 300 users were on just four VMs, and no one had complained.

The Coronavirus scare is causing a similar scenario – suddenly, more users are on the system working remotely than before. Their expectations have not changed. They just expect to be able to work. Well, that and have ten browser tabs with Facebook, WHO, CDC, and various news outlets all open at once. So, are you prepared? There are some 40,000 words I wrote on the topic, so go get the book, so I don’t have to repeat myself. The good news is that what you implement during this crisis will set your company up to save money long term. Worth it.

Quick Coronavirus Citrix Hero Wins Not in the Book

I wanted to highlight a few things that are NOT in the book that you should be aware of as there is an increased need for users to work from home. Here are a few time and crisis tested ways you can rise to this occasion.

Implement RemotePC

Up to this moment, my go-to event for recommending RemotePC was during the Toronto floods. While systems were largely intact, many people could not get to their offices. Sound familiar? Oh, how quickly we forget this feature built into Citrix Virtual Apps and Desktops! RemotePC allows the Citrix Hero to install a light VDA onto the users’ existing PC and add them to a Delivery Group. Just like that, the user has remote access to their PC.

I fully recognize a primary challenge here: The Desktop Workstation has been dead or dying for a long time. Most desks today have a docking station for a laptop or a thin client (or PC converted into a thin client thanks to my friends at IGEL). I do NOT recommend installing RemotePC onto a laptop. They tend to be configured to go to sleep or save power. This is not typically a great combination. That said, in a pinch – having a user leave their laptop, pushing the RemotePC VDA along with some Group Policy to control the power state, and then removing all of that after the crisis may be viable.

The key with RemotePC is twofold in a crisis like Coronavirus (COVID-19):

1. Provide Remote Access to company resources without additional risks, such as enabling VPNs. Because RemotePC utilizes your existing CVAD licensing, there are no added licensing or differences in external networking. This also means no added compliance requirements. It will, of course, consume a license – so keep that in mind.
2. Lessen the burden on VDI and Server Desktops and provide continued access to user-specific applications and configurations.

Prepare for Increased Citrix Gateway Bandwidth Use

I’m going to ‘fanboy’ for just a moment. I have always loved that you can increase bandwidth allowed to a NetScaler (sorry… Citrix ADC) with just a different license key. No ripping out the appliance or having to deploy a completely new virtual machine. But what I do anticipate is that many companies out there may double, triple, or further increase the amount of staff accessing remotely to apps, desktops, or RemotePC. Each of these connections uses bandwidth – even when it is as highly optimized as Citrix protocols are.

You will want to keep an eye on your ADCs or set up monitoring to alert you if bandwidth is consistently exceeding 70% or more of your licensed limit. In some cases, the bottleneck may be external bandwidth. Either way, I have a quick win tip: Reduce the Bandwidth consumed per session. There are a few different ways to accomplish this, but first, there is an important concept to make sure management understands, and that is that the experience must be less rich.

1. Reduce ICA Channels. Remember that each component configured that uses the ICA protocol will have a level of bandwidth assigned, and there is always a reserve, even if the components aren’t in active use. This means that if you can reduce the number of unused ICA channels, you can save bandwidth. Remember that these are user-context settings, by and large, so you can make reductions in just the users connected via Gateway with a policy. Here are a few that should be on the chopping block if you can work without them:
   • TWAIN redirection (what year is it?)
   • Port Redirection (it’s 2020)
   • USB Redirection (if not needed, which would be rare for remote access anyway)
   • Audio, especially microphone
   • Printing
   • Disk redirection & mapping
   • Clipboard redirection
2. Restrict ICA Channel Bandwidth. This is an often missed setting where you can set a limit of bandwidth (or combined with our next tip a percentage of the overall session bandwidth) for certain channels that need to stay active. This has two benefits.
   • Reduces overall bandwidth for features that provide business value such as clipboard redirection or local drive mapping.
   • Prevents spikes in usage from exceeding your Gateway or connection maximums, which can cause pauses for other sessions. For example, if you set Printing to use a maximum of 500 kbit/sec, you will cause the print job to take longer but will not impact the other sessions as a result. During a crisis, we all make sacrifices. An extra 15 seconds to print doesn’t seem like much to ask.
3. Restrict Session Bandwidth. Especially used with setting a percent of a session for channels, this setting can be great in constant-access scenarios. This setting is rarely used because to be effective, you must do a little bit of fortune-telling… you essentially need to predict how many active sessions will need bandwidth, padding against inactive, and all divided by the overall bandwidth available. This is not an exact science so don’t treat this as an authoritative formula… but a quick and dirty way to figure out your initial settings. For example:
   • 500 active sessions
   • 100 inactive sessions (connected, just not actively using the screen) estimated at 1kbit/sec to maintain
   • 100Mbit Connection/Licensed Speed
   • 100,000,000-(100×1,000=100,000)=99,900,000 bits/sec
   • 99,900,000/500= 199kbit/sec per connection.
   • No problem. But make that 5000 active sessions, and we have a completely different dynamic.
   • Your session bandwidth restriction for connections over the Gateway should not exceed the math above us.
4. The Big Win: Reduce Visual Transmission. This highly depends on the version of CVAD you are running, but there are a lot of options to reduce the bandwidth required to conduct this miracle we call remote display. All of these should be validated before simply pushing out via policy and very much keep in mind that they can be applied with additive filters (such as X user group when accessing thru a Gateway gets rule A, but Y user group in the same scenario may get rule B.)
   • Use Video Codec only for Actively Changing regions – this setting is time-proven.
   • If bandwidth is very constrained, set Do Not Use Video Codec. This is an “approach with caution” setting in my mind as it will often cause complaints. This setting will force Thinwire encoding, which can cause video to appear ‘choppy.’ Again – if we communicate with users that during the crisis they will have less capability, we can ready them to work differently.
   • Set a maximum framerate for the session. You can often reduce this to as low as 8 fps and still maintain a usable environment for task workers, but you may need 12 fps as a minimum. The point here is to set something, observe, and fine-tune.
   • Reduce color depth. Let’s face it, we all like the pretty, but going from 16-bit to 24-bit color was a huge jump. 32-bit color is a lot of information to have to send. I have, however, seen much less need to control this setting than in decades past. The codec is good at only sending what it has to. This is usually the last resort setting.
   • If you feel like going deeper into this, head over to the Citrix Tech Zone for a briefing.
5. Avoid Handshakes During this Time of Crisis. Go with UDP based protocols if you can. Enlightened Transport is brilliant. After all – remember that we aren’t supposed to be shaking hands to prevent the spread of Coronavirus, so let’s take the same attitude as much as we can with ICA, shall we? A few caveats here:
   • Using UDP doesn’t reduce bandwidth; in fact, it may increase it slightly. This changes in high-latency situations where TCP causes retransmits.
   • Make sure your firewall has 443 UDP and TCP open to the Citrix Gateway VIP
   • Make sure your Gateway is configured for UDP using DTLS.

Focus on Quick Wins, Citrix Hero

There’s much more I can say here that gets way more complicated and take more time.

Now is not the time to get it perfect. That time was last month before people started working from home more, hoarding toilet paper but missing the canned goods. #CitrixHero #Citrix #Coronavirus #COVID19 Click To Tweet

I watch Citrix Engineers and Admins go down rabbit trails all the time, chasing blog after blog of perfect recommended settings. If you have the time to fine-tune – great. But the book Be A Citrix Hero, as well as this article, reflect my philosophy when dealing with times like dealing with Coronavirus, or any for which you may not be fully prepared. For example, this is not the time to try and implement complicated settings like Browser Content Redirection – even though it may be perfect for reducing server loads during times like this. If you have the time and can test it well enough, fantastic. Just keep a few things in mind for any changes you’re making during these times:

1. Your Support Desk is probably already overwhelmed. Be their hero, not their villain.
2. Focus on the highest positive impact in the shortest time
3. Set expectations appropriately – Be Scotty, not _________ (insert politician full of empty promises here)
   • Don’t promise everything will be the same with a sudden increase in consumption. It probably won’t be; even if you take these steps, there will still be compromises to make.
   • Be extremely cautious, estimating timelines. Users that have not used remote access are now using it. This typically means a dramatic increase in handholding and trouble tickets.
4. Don’t skip testing. I know the timeline is immediate. But if you have to stop everything else you’re doing to validate changes – do that. You will ruin more than just your day in a scenario like this by putting out untested best intentions. I don’t care what any blog or book, including my own, say – TEST BEFORE YOU DEPLOY, HERO.

I hope this was helpful to you.

Now go forth without fear. I know that in times like this, the fear of the unknown often takes over from the real problems in front of us. This is your chance to lead. Confidence is just as contagious as… well, come on, don’t make me say it.

Now for the marketing thing I’ll say and twitch while I’m typing it:

Working from home? Perfect time to pick up a copy of Be A Citrix Hero! No one’s watching – go for it!

I know, I know.

I’m the worst.

But I wouldn’t say all this if I didn’t genuinely believe that this will help you!

Be safe out there, and for the love of all that is holy, wash your hands!