6.7 - What am I missing here?

[semery67]

Long time VROPs user here, since Cap IQ days.

Just recently did a fresh install of 6.7 for a new data center we are standing up.  I'm feeling pretty stupid as I realize there have been some changes, but it looks as though a TON of features we used to use are gone?  I'm hoping they are just hidden somewhere that I can't find them, but I'm staring to doubt that.  Specifically:

- Where are the analysis, stress and recommendation tabs and info?

- For policies - where do you now set things like CPU and memory over allocation polices and the like?

- For VM level troubleshooting and performance - what happened to things like oversize\undersized\reclaimable info?

Lastly, nearly every test VM we have stood up shows like 99% memory utilization.  I read another post here mentioning that might be a know issue, but wondered if there was any more info on that?

Bottom line.  Am I better off destroying this current build and going with 6.6 again?

This is very frustrating.

Thanks for any info.

[daphnissov]

There's a lot you're missing, it's true, because a lot has changed. You'd probably better read the docs and KBs on the 6.7 release. Bottom line: If you rely on some of that functionality, you may want to stick with 6.6.1 for now (or 6.5).

[semery67]

Thanks.  Been doing quite a bit of reading over the past two days, but just figured I was missing something\somewhere.

I was always a pretty hardcore VROPS zealot, but man, this is like 12 steps backward at this point.

[semery67]

Just following up.  After playing around again for to long, I blew the whole install away and did a 6.6.1 install.  Everything is back, perfect, and working as expected.

Here's hoping VMware brings back all those missing features soon.  I sent an e-mail to my TAM and sales rep regarding my feelings toward the new version.  I'm sure one voice can't do a whole lot, but considering our ELA is up for renewal, maybe they will listen just a little bit, lol.

[PSVM]

I have had to open 4 support calls.

One is for false positive security logon messages from Log Insight

One for the 100% CPU usage

One for Stress

One for reports that now miss a ton of useful information

For me this is the worst release of vROPS ever. I am not in a position to go back to 6.6 so will have to make the best of it. Glad we also have Veeam One as that is still reliable and does what it is supposed to do.

I am confused how the 100% CPU issue got through testing and why anyone would think it a good idea to remove so many of the useful features.

A huge backwards step

[silus]

We had to also revert back to 6.6 as well.

Take for example where in 6.6 you could go to a VM and see recommendations for how much cpu/memory to add/remove. They removed this and added it all under the capacity > reclaim dashboard. But there is no way to sort/filter/search for vms so if you have a lot of entries you can only browse the table manually and keep clicking next.

Absolutely awful awful awful design.

Not to mention most VMs now showing up as out of memory if they are caching data (as most modern OS do, and release that cached data if something else requires the space)

Whoever signed off this heap as complete?  They clearly rushed this out to meet the vsphere 6.7 release because it seems like it is in mid redevelopment where they've removed all the useful stuff, didn't have time to put it back in and just released it that way.

Anyone from Vmware care to enlighten us what the hell is going on here?

[TomOtto]

The CPU and memory recommendations are not missing, they're just under a different metric. There's a list of what's been removed/replaced from a metric point of view.

http://partnerweb.vmware.com/programs/vrops/DeprecatedContent.html

The Summary | Undersized/Oversized metrics show the same values as stress did previously.

[Iwan_Rahabok]

I've received feedback from 5 customers on this, but only 1 is active working with us (via the TAM). I hope that's you 🙂 If not, drop me an email at e1@vmware.com cc your TAM. I've drafted a workaround solution (which has been going thru internal review by TAM, PSO, etc) for customers who are unable to go back to prior version. So far the feedback has been very positive.

I'm based in Singapore, but spend 60% of my time traveling to see customers.

Thanks sunnydua201110141​ for sharing this bro.

[Iwan_Rahabok]

Q: Anyone from Vmware care to enlighten us what the hell is going on here?

A: I do. Email me at e1@vmware.com, cc your account team, and we take it from there 🙂 It's too long to explain the why and history, so best we talk over video conferencing (we use Zoom). I'm not officially from vR Ops team but have the privilege of working with both PM and R&D closely.

[daphnissov]

Iwan Rahabok​, I'd really recommend having this sort of Q&A discussion in a public forum where all can benefit from it rather than just a privately-held session behind closed doors. This release is very controversial and problematic for many customers and they would also like a chance at hearing the reasoning behind some of these decisions as well as how to move forward.

[Iwan_Rahabok]

Happy to discuss openly.

The reason why I offered a direct discussion is I don't have to be "marketingly-correct" 🙂

Plus, I received feedback from 5 customers so far. The feedback is not identical. Different customers use it differently.

See the attached doc for something I plan to post. I made it in editable, so you can comment on it. My email is e1@vmware.com.

I'm based in Singapore, which makes the time zone hard. I'm generally available 0830 - 2130 hours all day, unless I'm following wife's instruction to take her for holiday.

e1

+65 9119 9226

[silus]

Thanks for the insight.

Simpler is always better, but I think just not at the outright removal of core functionality if the replacement is not usable.

For example there should be somewhere on the vm object's page that says capacity recommendation. You should not have to browse to a Vm, see a problem, such a 0 days memory remaining, and then have nowhere to see what to do about it. You should not have to manually browse a seperate section to find this info. The analyse tab was great   Upon the removal of this, it would have been better to even have this info under 'Recommended actions' under the summary tab, to say what to add or remove.  or even better, have had this informatoin on the Capacity tab. Again it will say something like 0 days left, but no information about how much to add.

Also the two tables in Optimze performance > Overview for over and undersized VMs, they need to be sortable and searchable!  Should not have to scour tens of pages or more to find recommendations for a specific vm.

Unfortunately the workaround listed in the doc would not work for us, as we use standard licences everywhere, but it should help those on higher licences.  But hopefully this all gets pulled together in the next release.  It is just a shame it got forced out in such a state, 6.7 is absolutely not production ready if you have grown to rely on the core functionality 😕