Android device allegedly killing Exchange
Discussion
We've a hosted exchange solution for half a dozen users. The mobile service has gone down periodically, for all users. When this happens, attempts to connect with the MS Exchange connectivity checker show the activesync service returning 503 service unavailable, 401 unauthorised or just timing out. The host is blaming one user's Samsung Galaxy S5 - it's apparently generating excessive numbers of sync commands and this is resulting in mobile access for all of us going down. I've seen reports of this behaviour causing problems with earlier versions of Android and ActiveSync - tripping some throttling limit - but he's on 5.0. Their suggestion was to uninstall the email account, reboot and reinstall it, but it's apparently still doing it. We are told that none of their other customers have this issue...
My personal feeling is that if one misbehaving device is enough to take down your service for an entire client, you have a serious issue with your infrastructure, but despite escalating the issue to their senior management they are exhibiting slopey shoulders over this. We never had this sort of problem with our previous provider, though the user had a different phone back then. My other feeling is that the owner of the phone should chuck it in the nearest skip and get an iPhone
but that's probably not a constructive suggestion.
Anyone seen this before?
My personal feeling is that if one misbehaving device is enough to take down your service for an entire client, you have a serious issue with your infrastructure, but despite escalating the issue to their senior management they are exhibiting slopey shoulders over this. We never had this sort of problem with our previous provider, though the user had a different phone back then. My other feeling is that the owner of the phone should chuck it in the nearest skip and get an iPhone
but that's probably not a constructive suggestion.Anyone seen this before?
grumbledoak said:
Am I correct in my understanding that one Android client hits a throttling limit, the server act to protect the infrastructure, and you conclude "blame Microsoft"? And people wonder why IT is one step forward and one backwards each time. 
No, I don't blame Microsoft for one user's device denying service to our whole company. I blame, in so much as I blame anyone, our hosted exchange provider. We have had hosted exchange for years with another provider without this kind of problem, though I am willing to concede that this may be a device problem they didn't face.What I would expect to happen, if one user's device is causing a problem, is for that one user's device to be blocked. Not his and everybody else's. I suspect that the fault that the system is behaving that way is not Microsoft's. That is, if the outages really have anything at all to do with this device and are not an attempt to fob us off.
grumbledoak said:
Try fully turning off the Android device to get proper proof. If that cures it, leave it off.
Ah, yes. I'll tell our CEO that the answer to his mobile email sometimes going down is to turn his phone off. Nice one. The problem is intermittent, happens every couple of weeks, lasts a matter of hours. I'm not sure he's going to be willing to turn his phone off for a couple of weeks.
He is willing in principle to get a new phone, however I can tell that he is not remotely convinced that they aren't giving us the runaround. So my objective at the moment is to stop this device making excessive numbers of requests, get the provider to confirm that it isn't, and then see if the system remains unreliable.
There are millions of S5 handsets, and millions of Exchange servers. If this is genuine, I can't believe that ours is the only one in the world exhibiting this behaviour.
Hence asking here to see if anybody had come across this issue.
There are millions of S5 handsets, and millions of Exchange servers. If this is genuine, I can't believe that ours is the only one in the world exhibiting this behaviour.
Hence asking here to see if anybody had come across this issue.
hstewie said:
No, unlikely to get anything done server-side to fix it. If it was only affecting the one device then reboot and wait wouldn't be such a bad option, but it's taking everyone down.
They've said that they would typically expect 1000 hits / 24 hours, and that this device is at around 5500. Mostly sync requests.
They've said that they would typically expect 1000 hits / 24 hours, and that this device is at around 5500. Mostly sync requests.
Gassing Station | Computers, Gadgets & Stuff | Top of Page | What's New | My Stuff