4 ms·

What were the "hiccups"? That can't be good? P.S. what's your start-up do?

jjcm 14y ago
Disclaimer: I work for MS. The hiccups I've had have been DNS routing issues early on. It seems to have been solved, but for a while I was experiencing periodic enormous increases in latency (I think at one point the delay was 26 seconds). From what I heard they were specific to their linux VPSs (which is what I'm using), but those were solved around three or four months ago. I haven't had any issues since. As far as the VPS goes, overall it's decent. I do a lot of my development over ssh so small latency blips are almost always immediately noticeable. Linode seems to still be more consistent, but just barely.
sahaskatta 14y ago
We've had several issues over the past few months. To be fair, I should note that they did brand the Linux VMs as "preview" (which is just another word for beta). Even though they were already charging for the product, they didn't guarantee an up-time SLA either so we were well-aware it would be a risk using the product. The biggest issue by far? Random VM Reboots We thought our applications or some particular package was causing our VMs to randomly reboot. After thorough investigating, we found out that our code was perfectly fine. The folks on the Azure team were rebooting our VMs for internal upgrades. They oddly were doing this without any advance notice. This would affect our application servers, database servers, and even our development/staging instances. Another frustrating aspect was the fact that they would often do these upgrades during peak traffic hours (9 AM - 5 PM Pacific Time). Without a better solution, we just wrote some code to ensure that our application supervisor would auto-start if the system were to reboot which helps mitigate the problem. Hopefully these issues no longer occur now that they offer an 99.95% SLA. As for your other question, my start-up is Pepperdeck https://pepperdeck.com/ https://pepperdeck.com/

jjcm 14y ago

Disclaimer: I work for MS. The hiccups I've had have been DNS routing issues early on. It seems to have been solved, but for a while I was experiencing periodic enormous increases in latency (I think at one point the delay was 26 seconds). From what I heard they were specific to their linux VPSs (which is what I'm using), but those were solved around three or four months ago. I haven't had any issues since. As far as the VPS goes, overall it's decent. I do a lot of my development over ssh so small latency blips are almost always immediately noticeable. Linode seems to still be more consistent, but just barely.

sahaskatta 14y ago

We've had several issues over the past few months. To be fair, I should note that they did brand the Linux VMs as "preview" (which is just another word for beta). Even though they were already charging for the product, they didn't guarantee an up-time SLA either so we were well-aware it would be a risk using the product. The biggest issue by far? Random VM Reboots We thought our applications or some particular package was causing our VMs to randomly reboot. After thorough investigating, we found out that our code was perfectly fine. The folks on the Azure team were rebooting our VMs for internal upgrades. They oddly were doing this without any advance notice. This would affect our application servers, database servers, and even our development/staging instances. Another frustrating aspect was the fact that they would often do these upgrades during peak traffic hours (9 AM - 5 PM Pacific Time). Without a better solution, we just wrote some code to ensure that our application supervisor would auto-start if the system were to reboot which helps mitigate the problem. Hopefully these issues no longer occur now that they offer an 99.95% SLA. As for your other question, my start-up is Pepperdeck https://pepperdeck.com/ https://pepperdeck.com/