VIPRION, vCMP and appliance migration
Hi, I wonder what is best practice to calculate resource requirements of vCMP guest when migrating from standalone appliances. Let's say that we gave BIG-IP 2000S that with 60% of CPU load on average running LTM and ASM. What resources (number of vCPU) should be assigned to vCMP guest to at least get the same performance? 2000S is 1 Dual-core (+HT), one TMM process 1, 4 TMM threads (HTSplit disabled), 2 TMM threads (HTSplit enabled), 8GB RAM. To run both LTM and ASM I need at least 8GB RAM assigned to vGuest (wonder what this info from release notes really means "A vCMP guest provisioned with 8 GB of memory has less than 8 GB of memory actually available and thus does not fit in this category."?) For B2250 blade seems that at least 4 vCPU is necessary (2 vCPU gives around 6GB RAM). Is setting number of CPU to same or higher number (because of RAM) is enough to get same performance as appliance? Any other factors that are important? Piotr
MS Exchange 2013/2016 - Sizing questions - anyone with experience on SSL-TPS with 500K users?
Hello DevCentral community - I am looking for help sizing a MS Exchange environment with 500,000 users. The F5 would be performing LTM, offloading SSL with 2K certs. Application traffic is everything Exchange related - so, thick Outlook client, OWA traffic, and mail applications (POP, IMAP...). Does anyone know what we are talking for SSL-TPS w 2K certs for a user base that large? From this - I am trying to extrapolate the required equipment. (IE. a Viprion B2250 blade can handle 44000 SSL-TPS with MAX SSL license... is this enough? too much?) If anyone has experience for sizing, estimate figures, horror stories, please share. JS