July 2017 archive

The Schedule Builder for VMworld 2017 US opened today and registered attendees can start scheduling sessions. One thing I noticed was they removed the ability to see how many seats were left in a session which is unfortunate as it was a nice was to gauge how full a session was getting. Also as a speaker it’s nice to see how your session registration numbers are progressing.

While building your schedule you will notice a distinct lack of VVols sessions this year, you can read more about why that is here. I will make the pitch for my own VVols session which will be a technical deep dive on the new VVols array replication feature introduced vSphere 6.5. I have one of our 3PAR engineers who developed much of our integration joining me, so despite the session being a sponsor session I can ensure you that is will be very technical content and not marketing focused. I also have Julian Cates from Nimble joining in as well to tell us a bit about their implementation of VVols replication.

So I’d love to see you at our session and being it’s at the end of the day on Monday if you want to catch a beer afterwards and talk VVols I’d be happy to do that. Now if you are interested in VSAN sessions there is definitely no shortage of those, almost 1/4 of the sessions at VMworld are on VSAN. A search of VSAN in the breakout sessions in the Content Catalog show 89 sessions on VSAN of the 425 total breakout sessions. So go get registered before sessions fill up and hopefully I’ll see you there.

While we wait for the results announcement which should come in about 2 weeks via a live webinar with John Troyer and Turbonomic, I thought I would release some statistics on the results as an appetizer before the big meal arrives.

• This year there were 1600+ votes, which is about the same as last year.
• There were 321 blogs in the voting last year, this year there were 273. Blogs that did not have at least 10 posts in 2016 were left off the voting ballot this year.
• There were 62 new blogs added this year to the ballot that were not there last year.
• There was 1 new blog (started in 2016) that made the top 50.
• There were 11 additional blogs that made it into the top 50 that were not there last year.
• There were 8 blogs that made it into the top 25 that were not there last year.
• There was 4 new blogs to the top 10 (one blog from last year did not qualify and Duncan/Frank chose to step down).
• There was 9 position changes in the top 10.
• Do we have a new #1? Of course we do as Duncan did not participate, who will it be? Watch the live results show to find out.

So the voting closed about 2 weeks ago and we had over 1,600 people vote this year. Afterwards I had to export all the survey data to Excel and apply the weights to all the votes to get a total point score for each blog. The way that works is a #1 vote is worth 12 points, #2 vote is worth 11 points and so on down to a #12 vote is worth 1 point. In the past I’ve had to manually count the number of #1, #2, #3, etc. votes for each blog but starting last year I was able to automate it in Excel so it’s much simpler and accurate.

This year there are two additional point modifiers that are applied to each blog that factor into the total point score. These were added in an attempt to make this less about the popular vote and reflect the work the bloggers put into maintaining their blog. The first is the number of posts that a blog had in the year, I applied 2 points for each post up to a maximum of 400 points, so the more active a blogger is they are rewarded for it and will be ranked higher. For any site that was a pure blog aggregator with automated content they were scored at the minimum 10 posts.

The second score modifier is based on how well a site is constructed and how well the end user experience is. Any blog that was bloated and inefficient is scored lower than blogs that are optimized and fast. The measurement for this that I used was Google Pagespeed which analyzes a website and scores it with a percent with 100% being best. I ran every blog through Pagespeed and the resulting percent is used as a score modifier. Every blog started with 200 possible points available which was reduced by the resulting Google Pagespeed percent. So if your blog scored a 70% it would have a 140 point modifier applied to you your total point score.

So the resulting score looks like this:

• Voting points + post points (max 400) + Pagespeed points (max 200) = total point score

I’ve spent hours and hours in Excel getting all these modifiers in, thanks to Andreas’s blog post counts it helped me immensely not have to count blog posts for each blog but there were still a few dozen that I had to go count. So all the scoring is now complete and we’ll be announcing the results via a live Top vBlog Results Show hosted by Turbonomic with myself, Eric Wright and John Troyer. We’ll be using a standard webinar platform this time around as Google Hangouts was rather limited and a pain to use.

Also this year instead of doing commemorative coins we’re giving nice 1.5″ lapel pins to the Top 50 bloggers. They will be different medals just like the coins were for Top 10/25/50 and because there are no die costs we can have them made Top 10, Top 25 and Top 50 instead of all of them saying Top 50. The pins will be done before VMworld US and you can pick them up there, if you are not attending we can mail them to you.

So stay tuned for the details, the webinar will be held the first week of August.

Despite an initial rough start with the NetGear Orbi Home Wifi System I’ve been a big fan of it up until today. The reason I bought the Orbi was because I kept having wifi drops in some areas of my house with my Asus RT-AC87U router due to it’s placement. I play a lot of online games like Clash of Clans and if you know CoC the dreaded wifi symbol while you are playing is not something you want to see. That symbol means your connection sucks and you are about to get dropped, I would get that frequently with my Asus router at the far end of my house. Once I had Orbi up and running my wifi problems were completely gone for 7 months, until today.

Netgear pushed out a 1.12.0.18 update to my router and satellite last night and today I awoke to constant and frequent disconnects on both my iPad Air 2 and Galaxy S7 Edge. I could literally not go a minute without a disconnect. I called their support who was basically useless, they said to enable Implicit Beamforming which has been disabled by default many updates ago because of issues it caused with some devices. I tried that anyway and it seemed to help a little with the iPad but not the Galaxy S7.

Ironically Netgear had a big focus on fixing wifi disconnects with this update which has apparently plagued them from the beginning, read the Netgear forums and there are lots of threads on this issue like this one and this one. They also have problems with the Orbi losing connection to your internet device (i.e. cable modem), I also experience this frequently as well, the only fix is to reboot the cable modem. Most of the updates up until now have not really done all that much, this one had a lot of big fixes as seen below:

• Fixes the iOS disconnection issue.
• Fixes the Dropcam disconnection issue.
• Fixes the convergence issue when an Ethernet connected device moves from one Orbi to another.
• Fixes the issue where the Orbi app can’t find the Orbi router if the router is using IPv6 and is in AP mode.
• Fixes the issue where the Orbi app can’t display more than 20 connected devices.
• Fixes the issue where the installation assistant might not display if the device is connected to the Orbi network wirelessly.
• Fixes the issue where the 2.4 GHz backhaul setting is disabled if the fronthaul MU-MIMO and TxBF is disabled.
• Fixes the issue where IPv6 devices on the guest network can access the Orbi router’s web GUI.
• Includes security fixes for the following security vulnerabilities:PSV-2016-0133, PSV-2017-0607, PSV-2017-0615, PSV-2017-0736, PSV-2017-2190

So an update that supposedly fixes disconnect issues (which I never experienced) does the opposite and starts causing them instead. Thank you Netgear for breaking my perfectly wonderful wifi system. Apparently their engineering team has been spending most of their time introducing new Orbi devices and not spending much time refining their current firmware.

I ended up downgrading both my router and satellite to the prior version which is 1.11.0.20 which has been working perfectly. In fact I have never had an issue with an Orbi update until now, I’m beginning to suspect the reason why is that they haven’t really done much with any of their 11 updates up until this one. Look at their typical release notes for the last 3 updates:

• 1.10.1.2 – Fixed few bugs
• 1.9.1.2 – Fixes bugs and security issues. – Improves the firmware upgrade process. – Continuous improvement in wireless connection stability
• 1.8.0.6 – Fixes bugs and security issues. – Supports Korea wireless region

Now that they have finally starting to pay attention to some of the serious issues that Orbi has had they have apparently tried to fix some things and broke other things. At this stage with this new update I’m not going to update past 1.11.0.20 until Netgear gets their act together which is a shame as I will be missing out on potentially other fixes and enhancements. I loved my Orbi up until today and just yesterday was recommending it to people. However after this update dropped I will not recommend it anymore and may just end up migrating to one of the competing systems from Ubiquiti, Google and Eero which is a shame as Netgear had a good thing going with Orbi. Ultimately it comes down to what works best for me and if Orbi can no longer deliver that anymore than it’s adios amigo.