Azure Outage... Again
-
We aren't wondering if Azure is down, we know that it is. We know that the issue is Microsoft's and that it is the same issue that they have been having over and over again with many companies (most that we've talked to, actually, it's more than 50% of companies that we've interfaced with report that this exact issue is one that they have experienced and have experienced MS denying it - even to our faces.) What we are asking is how localized is it. Is it just one account (maybe an account manager deleted an account.) Is it regional. Is it people on a single database server or account category?
-
MS support responded much more quickly than they stated that they were likely to do and... they need our subscription info to process the ticket.
AAARRGGHH
-
Well I responded right away when they responded to the ticket. Not that I can give them any information
-
@Minion-Queen said in Azure Outage... Again:
Well I responded right away when they responded to the ticket. Not that I can give them any information
A lot of the issue that we have found in having this happen a lot, is that support is based in India and has a script to follow and, of course, if a script can handle it, they have already automated the fixes. So even though this problem comes up constantly they act like there is no solution and just freeze up. There have no process for dealing with these things, even when they are routine.
-
OK. So...
O365 with hosted Exchange - Good idea.
Azure - Bad idea. -
@wrx7m said in Azure Outage... Again:
OK. So...
O365 with hosted Exchange - Good idea.
Azure - Bad idea.YUP!
-
@wrx7m said in Azure Outage... Again:
OK. So...
O365 with hosted Exchange - Good idea.
Azure - Bad idea.That is not to say that O365 doesn't have it's trials and tribulations,.. But it has been the most stable.
-
@wrx7m said in Azure Outage... Again:
OK. So...
O365 with hosted Exchange - Good idea.
Azure - Bad idea.Consistently, that is what we see.
-
The important part of that discussion is when you need Exchange, then O365 is a good idea. If you don't demand Exchange, there are great alternatives that cost the same, or less, and are more reliable. But they aren't Exchange.
As long as you need Exchange, you are beholden to MS support and capabilities to a significant degree either way.
-
Why are all of Microsoft's failures colossal ones, such as the whole IMAP situation recently? They seem to take an inordinate amount of time to fix as well.
-
@tonyshowoff said in Azure Outage... Again:
Why are all of Microsoft's failures colossal ones, such as the whole IMAP situation recently? They seem to take an inordinate amount of time to fix as well.
Partially because they only make colossal systems.
-
@scottalanmiller said in Azure Outage... Again:
@tonyshowoff said in Azure Outage... Again:
Why are all of Microsoft's failures colossal ones, such as the whole IMAP situation recently? They seem to take an inordinate amount of time to fix as well.
Partially because they only make colossal systems.
Naturally, but it doesn't seem as though even larger systems have as much in the way of outages. Not to brag, we aren't nearly as big, but we haven't gone down once since 2007. I bet I've jinxed myself.
-
@tonyshowoff said in Azure Outage... Again:
@scottalanmiller said in Azure Outage... Again:
@tonyshowoff said in Azure Outage... Again:
Why are all of Microsoft's failures colossal ones, such as the whole IMAP situation recently? They seem to take an inordinate amount of time to fix as well.
Partially because they only make colossal systems.
Naturally, but it doesn't seem as though even larger systems have as much in the way of outages. Not to brag, we aren't nearly as big, but we haven't gone down once since 2007. I bet I've jinxed myself.
See it's this type of issue that SMB's look at and ask - why in the hell would I want to move to the cloud. In 20+ years and around 300 servers worth of support, I think I've lost 4. This is such a small chance of issue that the expense, albeit supposedly better quality setup, extra internet, etc all seem barely worthwhile in the face of outages like these.
I'm going to curse myself as well, but I've had Exchange for 5+ years now, and Domino for 12 years before that and never a full out failure other either hardware or software.
I had tons of clients in the same boat.
just shaking head - not sure what's real anymore
-
@tonyshowoff said in Azure Outage... Again:
@scottalanmiller said in Azure Outage... Again:
@tonyshowoff said in Azure Outage... Again:
Why are all of Microsoft's failures colossal ones, such as the whole IMAP situation recently? They seem to take an inordinate amount of time to fix as well.
Partially because they only make colossal systems.
Naturally, but it doesn't seem as though even larger systems have as much in the way of outages. Not to brag, we aren't nearly as big, but we haven't gone down once since 2007. I bet I've jinxed myself.
Oh granted, MS is not on top of these things like their competitors. They are "good" (mostly, not Azure, they couldn't cloud their way out of a plastic bag these days) but they are not "great." They clearly have zero capability to play with the Amazons and Googles of the world.
Partly, and this is huge, MS is a software firm, not an IT one. Running large scale IT is as foreign to them as it is to any random large scale company. And they are hampered by needing to run a lot of it on Windows and Hyper-V which are obviously not the best choices in many cases. Amazon and Google, for example, choose other platforms for a reason. Microsoft has an "eat their own dogfood" problem that limits their choices significantly and limits them to using technology that no other major player would even consider.
-
I'm going to curse myself as well, but I've had Exchange for 5+ years now, and Domino for 12 years before that and never a full out failure other either hardware or software.
This isn't normal, though. I've worked in many a shop with massive budgets and huge teams and their Exchange and similar systems failed far, far more often than Microsoft does.
And you are talking Exchange, we are not. We are talking Azure. MS Hosted Exchange I've barely seen blip, it's way more stable than any on premises / in house I've ever seen. It's Azure that they can't run to save themselves. If you can run a cloud the size of Azure in house without a failure, let me know.
-
It's weird... It seems the more complex your system is, multiple DAGs, and the other components, the more likely there will be problems. The other company I constantly mention, their Exchange infrastructure is (or at least a few years ago, was) was constantly crumbling.
-
@Dashrender said in Azure Outage... Again:
It's weird... It seems the more complex your system is, multiple DAGs, and the other components, the more likely there will be problems. The other company I constantly mention, their Exchange infrastructure is (or at least a few years ago, was) was constantly crumbling.
That's a HUGE factor in any HA discussion. HA systems bring their own risks. At some point they introduce more risk than they mitigate.
-
@Dashrender said in Azure Outage... Again:
It's weird... It seems the more complex your system is, multiple DAGs, and the other components, the more likely there will be problems. The other company I constantly mention, their Exchange infrastructure is (or at least a few years ago, was) was constantly crumbling.
This is a common view point, the more complex a system is, the harder it is to manage and the more likely there will be a failure. The Saturn V rocket was so complex they could expect at least 600+ parts to fail during any launch, or maybe it was 6,000, well it was a lot and that is pretty frightening. That's why we need to keep things as simple as possible. However, there is a train off, simple systems don't scale worth a damn. Keep it as simple as you can with what you have.
-
@tonyshowoff said in Azure Outage... Again:
@Dashrender said in Azure Outage... Again:
It's weird... It seems the more complex your system is, multiple DAGs, and the other components, the more likely there will be problems. The other company I constantly mention, their Exchange infrastructure is (or at least a few years ago, was) was constantly crumbling.
This is a common view point, the more complex a system is, the harder it is to manage and the more likely there will be a failure. The Saturn V rocket was so complex they could expect at least 600+ parts to fail during any launch, or maybe it was 6,000, well it was a lot and that is pretty frightening. That's why we need to keep things as simple as possible. However, there is a train off, simple systems don't scale worth a damn. Keep it as simple as you can with what you have.
The Saturn V was so complex that even NASA and the USAF lack the knowledge, ability and information to build them today. Yes, the US has actually moved backwards in rocket technology (for that size and type of rocket) and can't do today what we could do forty years ago. A bit crazy. It's one of the biggest "lost knowledge" items in the last century.
Much like how the Romans knew more about making hydrolic cement two thousand years ago than we know today.
-
@scottalanmiller said in Azure Outage... Again:
Much like how the Romans knew more about making hydrolic cement two thousand years ago than we know today.
I agree with you, but I try to avoid stating things in that way, because it helps reinforce that "ancient knowledge" garbage that's out there, like "they could do things we couldn't do, therefore here's an insane conclusion about how they had magic/aliens/electricity/etc or were better than us and we don't know anything, p.s. here's more new age garbage." The same thing happens with the Pyramids as well, and it's often not that we can't do it, we just don't know how they did it, but we may figure out eventually.
Additionally, their approaches were often radically different, again the Pyramids, we couldn't build them that way today not just because we're not absolutely sure how they did, but even if we knew, dealing with concrete is easier than giant stones, so our building infrastructure is based around that, as opposed to oxen pulling sleds or whatever. Going way back to the 60s it was claimed "our cranes cannot lift the stones in Egypt, therefore aliens," of course they didn't use our cranes so that's not even important.
That was slightly off topic, but I think it does also show another lesson about how approaches can be so different that building something with them, even in software or whatever, cannot be done the same way just because it's all so functionally different from the bottom up.
