There isn’t any. Tags the crux of the whole thing.
1 Like
Both App and Infra are provider responsibilities. What part of ‘we changed app and didn’t test on and cannot even try to repo on our own infrastructure’ is customer responsibility?
1 Like
Well, in the US we have these things called rolling blackouts where the utilities unplug whole grids to save power and then move on to another grid for the next hour.
In other areas if too much electricity is used, everyone just “browns out”. The lights dim and devices that have bigger draws just stop working. It is exactly the same as the noisy-neighbor problem we see in Azure.
You are right though, infrastructure issues (configuration, monitoring, hardware, printing etc.) are the responsibility of the provider. As mentioned above, I agree with Josh that there should be a short-cut to the cloud team.
Every company is different. When we had to have follow-the-sun support for three continents, it was a no-brainer. Epicor Single Tenant was cheaper than hiring operators for three shifts. It made our SOX audit cheaper since direct access to the database wasn’t available. For other companies with little or no technical expertise, it may also make sense. I’m in the camp that some things are better in the cloud (web hosting, email, payroll, …) while other things are better on-prem (CAD, IIoT, metrology, …).
2 Likes
Ouch, if I didn’t have generators that would want to make me move the servers to somewhere with a bit more stable powersupply.
1 Like
Not everywhere in the US. We’ve never had rolling blackouts in the Midwest where I am. The power companies here invest in their infrastructure and manage their “network” better than other power companies in say… California.
It’s not an Azure problem though. It’s an Epicor problem. They poorly manage their Azure resources. When Jose is pounding on his Azure tenant it doesn’t impact me. But when one Epicor SaaS tenant gets wild it impacts everyone. It shouldn’t if it was built correctly.
2 Likes
Texas has had rolling blackouts much more recently than California. But you are correct, we’ve not had any rolling blackouts here in Wisconsin either nor when I grew up out west in Utah. A few blackouts here and there because of storms, idiot driving into a pole and downing the line, etc though.
No, we just lose the entire Northeast for several days. Here in SE Michigan, I can count on three to four days without power per year because of trees falling during wind or ice storms. For whatever reason, Detroit Edison will not bury electrical lines. 
Only one of my customers had full power backup (medical lab) where I could work. Fortunately, it was a Friday and my on-prem customers were down until Sunday.
Well, no. You’re on-prem.
The problem with this (and many other posts here) is that we say cloud to mean specific and general things at the same time. That’s causing this post to bounce back and forth. All cloud, Epicor SaaS, Azure, … It’s difficult to have a conversation like this.
If we’re talking about Epicor SaaS, which I believe most of us are, then yes. It is not going well, and I remember when it was better. Epicor is definitely going through growing pains. Moving an on-prem Windows solution to Linux containers at scale is not easy, but it is the right thing to do. Kinetic is only one of many irons that Epicor has in the fire. (
) With everything going on at once, it has to be difficult keeping everyone rowing in the same direction. If every team was, SaaS would be on a better footing, and I think more people would move to it voluntarily and not have to be forced into it.
2 Likes
I was referring to workloads we do have in Azure.
I’m only talking about Epior SaaS.
Just giving you 
2 Likes
But its epicors decision when do do these things. As I stated before at no point should pilot have been changed w/o letting people either opt in or heck just tell them and absolutely shouldn’t have been done this close to a big drop like 2025.2. Going to beat the on prem drum again but I don’t have to put up with this sht. I choose as an when and co-ordinate with teams using the test or pilot system. Epicor is your supplier and your the customer what do they say … “the customer is always right”
Docker FTW though \0/ Would love some of that on-prem.
2 Likes
For their bottom line maybe, but not overall.
This should have been re-architected, or at least refactored before trying to shoehorn this stuff in. Especially on the timelines they are trying to push.
I’m sorry, but I don’t have the confidence in Epicor as a company I once did.
I’m no longer sure they are up to this challenge.
9 Likes
^ this
.net (not the framework one) has been container friendly for some time so it’s not that hard to migrated depending on how they chose to integrate with IIS. I suspect the thought of removing the Windows licence fee and the complexity of not using Kestral & IIS helped. Give it another 10 years and the db will probably be PostgreSQL too.
It won’t even be 10 years. I bet in the next couple years they jump to Postgress.
3 Likes
This along with docker is even more reason for an on prem version 
All driven by cost savings that no customer is going to benefit from.
2 Likes
I have to be careful what I say or I’ll be accused of starting a
Hopefully they in the end offer a container solution for those of us on-prem.
But the truth is. In return for Epicor saving money on Windows and SQL licensing we actually pay more in maintenance every year.
3 Likes