A short post to let you know about some upcoming speaking engagements that I am doing over the next couple of weeks.
First up, I will be speaking at the TechUG, or Technology User Group event next week. This event will be held on Thursday, November 26th. It will be held in the Westin Hotel in the heart of Dublin city, Ireland. There is a really good agenda for this event (which is not a VMware centric event), that you can find at this link here. I personally will be speaking about Virtual SAN (VSAN), VMware’s hyper-converged compute and storage platform. This will be more of an introductory type session, but I’ll also be giving an overview of new and upcoming features and where we are thinking about going next with VSAN. You can find the Dublin TechUG registration link here.
My next session is at the VMUGDK Usercon or Nordics Usercon, which will be held on Tuesday, December 1st. This event will take place at the Scandic Hotel in Copenhagen, Denmark. This year I will return to my roots and talk about core vSphere storage enhancements over the past few releases, and also a look at some upcoming plans. No VSAN, VVol or anything like that – this will be a discussion on VMFS, NFS, VAAI, PSA, etc. The Nordic UserCon details can be found at this link here. The registration link is at the same location.
If you are in the Dublin or Copenhagen area for any of these events, I’d love to see you there. I plan to spend most of the day at both events, so if there are any VSAN or vSphere storage questions or feedback that you’d like to give me, I’d be delighted to talk with you in person.
This is a new feature in vSphere 6.0 that I only recently became aware of. Prior to vSphere 6.0, all the I/Os from a given virtual machine to a particular device would share a single I/O queue. This would result in all the I/Os from the VM (boot VMDK, data VMDK, snapshot delta) queued into a single per-VM, per-device queue. This caused I/Os from different VMDKs interfere with each other and could actually hurt fairness.
For example, if a VMDK was used by a database, and this database issued a lot of I/O, this could compete with I/Os from the boot-disk. This in turn could make it appear that the VM (Guest OS) is running slowly.
I took the opportunity last week (while I was over in the Boston area) to catch up with Scott Davis. I’ve known Scott a long time, as he had various roles at VMware over a number of years. Scott is currently CTO at Infinio, a company that has developed an I/O acceleration product for virtual machines. The new version of Infinio Accelerator 2.0 released only a few weeks back, so I decided to reach out to Scott and find out about the enhancements that went into this new version.
A few weeks, my good pal Cody Hosterman over at Pure Storage was experimenting with VAAI and discovered that he could successfully UNMAP blocks (reclaim) directly from a Guest OS in vSphere 6.0. VAAI are the vSphere APIs for Array Integration. Cody wrote about his findings here. Effectively, if you have deleted files within a Guest OS, and your VM is thinly provisioned, you can tell the array through this VAAI primitive that you are no longer using these blocks. This allows the array to reclaim them for other uses. I know a lot of you have been waiting for this functionality for some time. However Cody had a bunch of questions and reached out to me to see if I could provide some answers. After conversing with a number of engineers and product managers here at VMware, here are some of the answers to the questions that Cody asked.
In Virtual SAN version 6.0, VMware introduced support for an all-flash VSAN. In other words, both the caching layer and the capacity layer could be made up of flash-based devices such as SSDs. However, the mechanism for marking some flash devices as being designated for the capacity layer, while leaving other flash devices as designated for the caching layer, is not at all intuitive at first glance. For that reason, I’ve included some steps here on how to do it.
In vSphere 6.0, an improvement has been made to how we handle I/O issues, such as flaky drivers, misbehaving firmware, dropped frames, fabric disruption, dodgy array firmware, and so on which can cause I/O failures. The issue is that, previously, we continually retry these sorts of I/O errors, which can lead to all sorts of additional problems. In this release we are changing our behaviour for marking a path dead.