According to Google's own calculations (in the email they sent about the price changes), this will increase our GCS bill by about 400% (and our entire Google Cloud bill by about 60%).
It would seem that we have until October to move elsewhere... :(
> It would seem that we have until October to move elsewhere
the biggest fear especially with this class of infrastructure (long term cold storage) is that they can make it too expensive to leave at any time by upping the retrieval / egress costs. How expensive is that move going to be?
Not sure about your usage model, but if you have a defined retention for customer data (e.g. 1 year), then you can start pushing new data to a different tier/cloud provider, as the old data drops off, and a year later, you're completely "migrated", without having to stage a stop the world migration between services.
In our case, it's effectively "forever" storage. Which is to say, we are obliged to retain data for at least 10 years but targeting 20 years.
At this point we are doing this with on-premise tape backup but that is in part because I'm yet to be convinced that we can trust cloud providers with this, especially since our future retrieval needs are unclear (some outlier scenarios could see us needing to retrieve substantial fractions of the data). Not to mention that even the coldest cloud storage seems to still be substantially more expensive than DIY tape archival (admittedly, not taking into account things like internal IT staff costs etc).
You're expressing a preference for convenience over security. The truth is that most people pick bad passwords, and even good passwords can be cracked.
2FA with a physical component is generally the best way to achieve the goal of "information flows with me". With a password only, you can more aptly describe the situation as "information flows with anyone who knows my password".
In that case, can we do 2FA with something biometric? Or even 2 passwords?
A physical component has a lot of issues:
* It can be stolen or robbed at gunpoint. Torture, drugging, and hypnosis aside, your mind is much more secure.
* It can run out of batteries.
* It's one more thing you can lose. It's already annoying enough to have to remember to carry 7 or 8 things every day, including a phone, bike light, smart watch, tablet, battery pack, reusable utensils, and so on. I don't want to have to add more things to this list.
* It can be damaged by the elements.
* It can be difficult to give access to others who you want to give access to.
* It may have security holes of its own, both in hardware and in software.
* When damaged or robbed, the user is highly inconvenienced, to the point that they are unable to access their own money/accounts/etc. How do get food, water, and get home from the middle of nowhere after your wallet and phone have been taken from your person? With password-only methods, you could theoretically find a nearby public terminal, log in with a simple username and password, and get an ride/call a friend/file a report/do whatever you need to do.
* If it relies on cellular service, it may not work internationally if the user changes SIM cards or devices. For many that live near border towns and cross borders every day for work, this becomes a massive inconvenience.
Biometrics make great usernames but poor passwords since they can't be changed. Imagine a fingerprint system of some kind - someone images your fingerprint from, say, a leftover coffee cup (not hard or expensive to do), and you're pwned.
The Yubikey does not run on batteries. It requires no cellular service. It can be damaged by the elements but not easily. Most electronics would break before it does. Of course you can lose it, but you can lose anything. Attach it to something you care about, such as your regular keychain. If you want to give access to someone, register a second key and lend that key to them. Then revoke when they don't need it.
We measure web traffic with Google Analytics. Free, industry standard, works.
We measure app metrics and server/service health with Librato Metrics. A local statsd instance collects UDP packets from all over our infrastructure then consolidates and dumps them out to Librato every 60 seconds. We track thousands of data points every minute and it keeps up without issue. Librato offers great and flexible visualization of your data without breaking the bank - highly recommended.
Of course you can make something similar if you're a) willing to build it and 2) don't care about the tremendous amount of complexity you're adding to your application in order to do so.
Both of those sound like bad deals if there's a packaged solution like Datomic that's built specifically for the use case.
According to Google's own calculations (in the email they sent about the price changes), this will increase our GCS bill by about 400% (and our entire Google Cloud bill by about 60%).
It would seem that we have until October to move elsewhere... :(