SAP on AWS Innovation News: April 2020

Posted 08 April 2020 by Ben Lingwood

Welcome to this quarter’s edition of our SAP on AWS Innovation Newsletter, our regular update detailing the new features, capabilities and market updates that will help drive new innovations and optimisations into SAP-Centric Workloads on AWS.

Reading from the script of a big-screen movie, like me I’m sure many of you are feeling quite bewildered by the recent global events, conflicting statistics pouring from the daily news channels, personal concerns about our own family and friends and even the most acclimatised “Homeworking” people are being isolation stress tested. Whilst we all keep our heads down and wait for the storm to pass over, there have been some notable technology updates released over the last quarter. Here are the best we have leveraged from the last three months in our landscapes;

  • Starting with COVID-19 related topics, AWS are offering FREE Workspaces and WorkDoc tiers, in addition, to access to AWS Chime for remote working. We use Chime extensively at Lemongrass. It is by far the most reliable conferencing tool we’ve experienced for multi-country video conferences for up to 50 users. See our previous post on a recent successful rapid deployment of 250+ desktops over a weekend.
  • New AWS Backup models have been GA released over the last quarter in all key regions enabling an entire EC2 instance to be backed up in one job. In addition, the EC2 backup images can now be replicated to second region(s) as part of the native functionality. This is also executed with a specific IAM role, which means it doesn’t short circuit the account compromise model. Finally, restoring just got a whole load easier as the binary format update means a single file can now be restored from the backup on EFS vs the complexity of a full filesystem which was needed before. We are in the process of building this new capability across our management platform. It is also great news as it removes another dependency from a third party we used in this space which reduces costs further.
  • If you missed the sales in January and February, you might have missed Cloud Endure had an 80% price reduction! We have used Cloud Endure for migration, backups and DR models in the past and it is especially good for non-SAP application migrations and cross-region backup images as only the data blocks that have changed are replicated (vs copying over full backup images) driving further ingress/egress charge savings between regions. If you’re using this for a DR scenario, remember it is only supported with AWS as the DR target for SAP.
  • A handy one for anyone using persistent Spot EC2 instances for training, development or DevOp scenarios, Spot instances that are backed up on EBS can now be stopped and started at will avoiding the usual interruption issues. Essentially the session is restored from the EBS image to the last recorded state. This assurance unlocks access to the much cheaper spot pricing without the traditional risks that usually prevents this EC2 model being used so much in the SAP space.
  • GuardDuty has an improvement release. The quality of threat detection can reduce false positives by up to 50% across the key monitoring areas such as SSH brute force, Port probes, DNS exfiltration, etc. This helps keep traffic down and reduces the cost of the alerting framework (SNS, etc.) As heavy GuardDuty users in our managed accounts, we’ve certainly noticed the improvement in alerting triggers. Also, GuardDuty had a reasonable price reduction across Flow Log & DNS Log analysis in the last week which again drives improved landscape operational costs down month on month.
  • We recently saw SAP certify some additional EC2 instances for SAP use in the Master Certification Guide. The r5.8xlarge & r5.16xlarge systems now certified for both OLTP and OLAP workloads. This nicely fills a large gap that historically existed between the 384GB and 768GB systems offering a 256 / 512GB VM image option. So, if you are one of those people with systems that fell between this gap, I’d recommend you run a quick scan / trusted advisor check so you can take advantage of this. Or, run some housekeeping to see if you can right-size your DB down to save some operational costs.
  • There are a few heads up points on the SAP front: SAP has released an update on HANA 2.0 SPS05 and the re-alignment of what that means with the maintenance schedule. One “Gotcha” to be aware of, which is hidden in the small print, is that, after 2025, SAP will only support the latest 3 EHP’s. So if you are an >N+3 type of organisation, keep an eye on your future patching strategy or take a look at deploying Near Zero Downtime Patching methods to help keep inside the supported range.
  • In the last few weeks, Lambda has also moved over to the Savings Plan model which can be called off against a 1 or 3-year commitment. This puts a new resolution on some of the serverless executions that are used around most SAP Landscapes providing access to a 17% cost saving for Lambda executions which adds up over time.
  • On the topic of cost monitoring, AWS Cost Explorer now offers Savings Plans recommendations for Member linked accounts. Historically, this was an issue that made complex multi-account landscapes hard to analyse. This new update scans across linked accounts to understand what further savings are recommended.

In closing, with increasingly challenging metrics about its impact hitting our desktops every 12 hours and over a quarter of the planet in some form of lockdown, you might be wondering what else can we do to help? As it turns out, we techies can do something. If like me, you were an avid SETI@Home distributed networking adopter back in the day, you’ll know how powerful socially distributed grid computing can be. The brains at Stanford School of Medicine have tapped into this concept once again with the “Folding at home” (labelled as such due to the way human proteins fold inside cells) project team who have adapted their grid project to tackle the problem of mapping the countless protein sequencing simulations needed to develop a vaccination. It’s a bit like buying a lottery ticket but every CPU cycle helps. If you have a spare vCPU, thread or GPU sitting around, it can be put to good use. Set up a team or feel free to join the Lemongrass team effort – 252167.

Keep safe, stay at home, and keep on innovating!

Ben
M: +44 (0) 7983 446723 | E: blingwood@lemongrassconsulting.com