halbritt

38 ms·

451.
by halbritt 8y ago
I tend to agree. When Yamazaki 12 was around $50, I'd go in for a bottle every now and then. At twice the price now, it's just overpriced relative to comparable Whisky from Scotland. I'd rather have a bottle of Springbank.
452.
by halbritt 8y ago
Here's a list of things from January of this year: - when scaling nodes, didn't get the desired count - creating and deleting clusters takes a while (30 minutes) - cluster operations get stuck in a loop - disappearing no
453.
by halbritt 8y ago
Development abilities would entail having the ability to take a task, or feature, write code around that task or feature including tests, and then submit a PR. As someone doing the hiring for such things, people looking to get hired seem to
454.
by halbritt 8y ago
"On the cloud providers it is fully managed so nothing really to troubleshoot" Clearly, you haven't run much at scale or experienced many problems. Troubleshoot a network issue through an ingress, kube-proxy and then down to
455.
by halbritt 8y ago
No disagreement here. There are, however, a number of people in the field that are all ops and have no development abilities. There is less and less of a need for those kinds of people. This has been hashed about at length, but the inverse
456.
by halbritt 8y ago
Indeed. That's why Google has SRE and Facebook has Production Engineering, and so on.
457.
by halbritt 8y ago
Infrastructure as software vs infrastructure as code.
458.
by halbritt 8y ago
Easy peasy. Now try troubleshooting it.
459.
by halbritt 8y ago
Support seems pathologically bad. Obviously, the support people aren't well trained in all the various products and modes of failure, but they make no effort, take no ownership, and don't really escalate to engineering staff with
460.
by halbritt 8y ago
I've used Azure a bit and it's been pretty painful for the most part. There are three, maybe four identity management systems (Azure native, MS Live, Hosted AD, others?). Working with the SDKs and doing programmatic authenticati
461.
by halbritt 8y ago
Sure. It's an extraordinarily complex system that's difficult to engineer correctly. It provides extraordinary durability, but the radius of failure isn't obvious. Pro tip, it's the entire cluster. As such, an issue
462.
by halbritt 8y ago
I lol'd while nodding vehemently in agreement.
463.
by halbritt 8y ago
OpenWRT is pretty powerful, certainly. If you prefer an open source solution, that's definitely the way to go. I've used OpenWRT which I liked quite a lot, but I prefer the ease of having an integrated solution with the Unifi pr
464.
by halbritt 8y ago
I think so.
465.
by halbritt 8y ago
Then set the liveness check for something further upstream of the ISP. Getting an AS is easy. Getting portable IPv4 address space or an LOA to readvertise is more tricky.
466.
by halbritt 8y ago
It's great in that it integrates with the unifi app and provides complete visibility. However, I've had a couple of the USG fail under certain circumstances with enough bandwidth (>200mbps) or tickling a certain feature of the
467.
by halbritt 8y ago
Reiterating this: Management and updates are wonderful through the unifi app. It functions similarly to Meraki, better in my experience and there's no recurring fees. The app can be self-hosted, run in the cloud, or on something they
468.
by halbritt 8y ago
Nice setup, but we can all pretty much agree it's overkill for most. My ISP is fairly reliable and outside of infant death, most network elements have a pretty long MTBF. I run a similar set of WiFi gear. I've a couple PoE power
469.
by halbritt 8y ago
I love Unifi. They're the only consumer-grade access points that will do roaming worth a damn. They're access points, though, and don't function as a router. I use the UAP-AC and an EdgeRouter. The EdgeRouter has relativel
470.
by halbritt 8y ago
> I love Ceph so much... Clearly hasn't been bitten by it, yet. I mean... I love Ceph, too, but I don't ever want to run it again.
471.
by halbritt 8y ago
IIRC, the Unifi stuff as well as Meraki will do multiple ISPs. They do outbound NAT, and have a liveness check which is just a ping sent to the next hop. Ping fails, or the interface goes down and the device simply sends the traffic the o
472.
by halbritt 8y ago
WOOP WOOP
473.
by halbritt 8y ago
The binary approach is that it either isn't considered or people opt in for all of it without consideration for what is actually needed. The Google SRE book goes into this at length. For each service, they define SLOs and make a cons
474.
by halbritt 8y ago
The correct DNS solution is to use multiple providers. See: Route53 and Dyn outages in the past couple years.
475.
by halbritt 8y ago
Agreed for the most part. Availability for very many is a binary operation. They either do none of it or all of it. A clever man once said, "you own your availability". An exercise in BC planning can really pay off. If infra is
476.
by halbritt 8y ago
Sight Machine | Infrastructure Engineer- Infrastructure Enginnering | San Francisco, CA | Onsite Sight Machine is a SaaS company that operates an IIoT data pipeline and analytics platform for manufacturers. Infrastructure Engineering develo
477.
by halbritt 8y ago
Most folks overestimate their need for availability and lack a willingness to accept risk. There are distinct benefits that come with avoiding "HA" setups. Namely simplicity and speed.
478.
by halbritt 8y ago
A few ansible scripts? Nah. Everything I have runs in kubernetes. The only difficulty I have to deal with is figuring out how to deploy a kubernetes cluster in each provider. From there, I write a single piece of orchestration that will d
479.
by halbritt 8y ago
"If you can afford it" There's a degree of complexity that comes with multi-cloud that's ill-suited for most early stage companies. Especially in the age of "serverless" that has folks thinking they don't
480.
by halbritt 8y ago
Ansible is great a doing the things that Ansible does!

More ›