The third episode of TalkCrunch is all about Amazon’s game changing new S3 Grid Storage web service that I wrote about on TechCrunch last week when it launched. Co hosted with Keith Teare and Nik Cubrilovic, we bring on two guests from the Amazon Web Services group: Adam Selipsky (VP of Product Management and Developer Relations for Amazon Web Services) and Dave Barth (Product Manager for S3).
In the first segment Adam gives us a high level view of all of Amazon’s web services. In the second segment, David gets into detail on the new grid storage service itself, including an in depth review of features.
Keith and I also try, without success, to talk Nik into building his new online storage company Omnidrive on top of S3. It’s a lively conversation, to say the least.
Other companies mentioned include PayPal, Jigsaw, Plaxo, Yahoo, Google and Riya.
Running time: 51:25m
Download Size: 23.5MB
This was such an informative and engaging podcast! A fluff piece??…softball questions??…none of that here. Thank you!
Keith’s comments…echoed by Dave Barth…regarding the difficulty of comparing “apples to apples” when evaluating hosting providers’ pricing platforms was of particular interest to me (and many other listeners?).
Any chance that some follow up on Tech/Talk Crunch will be forthcoming to help clear away the fog and provide folks who are building applications a better understanding of what type of hosting model would work best for them at various stages of their venture (i.e. comparing hosting solutions based on bandwidth scaling needs…and IT man hour needs on the back end…as the venture scales)?
Nik,
Cant wait to test your service when it comes out. I must admit though, I would have thought it would be smarter to use Amazon for storage so you can concentrate on your speciality. However I am obviously not privy to if you have other pressures like VC’s expectations, staff you have already hired to solve this solution… etc.
Regards
Clay
Co-Founder - Minti
Really enjoying these podcasts Michael. Fantastic stuff.
Please can you supply a .wma download for podcasts ?
[...] Episode 3 of TalkCrunch is up. We invited in two key Amazon S3 team members, Adam Selipsky (VP of Product Management and Developer Relations for Amazon Web Services) and Dave Barth (Product Manager for S3), to talk about Amazon’s exciting new Grid Storage web service. [...]
Pretty good overall, nice to hear the Alexa guy speak he sounds like a real sharp guy.
The guy with the English (?) voice talked too much about bandwidth costs without any real understanding and it got boring and embarrasing - found myself cringeing a little (sorry). If you’re going to ask detailed specific at least make sure you find out your facts first….
Rated B.
Is webservie fit for uploding / downloading very large files (say 500 MB movies). We have been working on this for all most 4 years (storge and reterival of large amount of large files) and WS is not the answer.
Jay
CTO, http://www.objectcube.com
It seems to me that the Omnidrive could still be a very serious competitor to S3, although the creation teams of the first and the other can not be compared. So the ultimate question will be, can the small ant bite a piece of pie from a bees possession.
Tim,
qubitsearch.blogspot.com
[...] Episode 3 of TalkCrunch [...]
Jay raises a good point as well - SOAP is slow and web services aren’t the most efficient way to move files around
Nik/Jay: The WS is not a good option for sure (unless you don’t slice the big file in pieces and xfer them one by one — but again won’t be very effective). For such big files probably resume-after-crash FTP will work much better.
We are using WebDAV to distribute all the files in our CDN, but we are not very happy with it (too slow, too much overhead).
Lenkov
SiteKreator.com
Good show. My only complaint so far is the intro music… Is there anyway you can change that so it doesn’t sound like the trailer for 2 Fast 2 Furious? Otherwise I’m enjoying it.
I could have used a bit more info on the exact goals/uses for the tech. I’ve read through the S3 website and I’m still stumped as to what the possible applications are. If it’s simple web storage for large files (like my companies video game download) then the price point isn’t actually all that great. We’re getting an average of .07 a gig transfered between our ISP and GoDaddy.
I guess I want to better understand the rudimentary aspects of what S3 IS which you seemed to skip… as did their site! LOL
OH! Wanted also to say I like the podcasts ALOT! They’re very informative even if this last one was over my head a bit. As far as the music goes, I don’t mind it but it needs to have some EQ ballance work done. I cranked down the volume because of the music and then had to crank it back up again because I couldn’t hear the voices.
I agree.
Although S3 seems very promising, the service lacks clear specs about performances, features (data redundancy and latency in particular) and API. I guess it will take a couple of months before we get a clear view of their roadmap and what startup could really do with this.
A very great thing with S3 is that they have provided examples in many language. Whether you are a pure Java programmer, a PHP hobbyist or a Ruby junkie, there are code examples available.
Ted
—
Notebook
Batteries
I was fairly scathing about Episode 2 but 3 was altogether better. Audio quality much improved and the show was more structured. The presentation was a shade more engaging.
The subject matter was interesting and topical, and the discussion only slightly overlong. It’s getting there.
Do please sort out the intro music. It launches in far too loud, we don’t know what we’re listening to for quite a few seconds until the voice cuts in then there is still a disconcerting sudden volume drop.
Being audio only the voice has to start from the very beginning to inform the listener what they’re listening to, with the music in the background or cutting in after, and volume levels kept constant. Maybe when your signature tune is instantly recognisable world wide you can leave it to play on its own for a bit.
I found the conversation very interesting, but the whole piece about the cost of bandwidth went a little off-track. Its easy enough to provide a load of disks and a lightweight front end, but to really deliver you are going to need a multi-tiered architecture with data replicated at different data centres and redundant internet backbone connections. Amazon already have a lot of this infrastructure in place of course.
The monthly standing charge per GB is obviously going to cover the storage, dev work and administration to provide the solution, but the bandwidth charge needs to cover the users share of the activity on the servers (and on replicas) - not just the data flowing in the wire.
Thanks for segmenting!
Greetings,
Manfred
[...] Talk Crunch Podcast on S3 [...]
I agree with Nik.
Omnidrive’s choices (as I see them):
a) A back end that they have already developed, that they control, does exactly what they want and which they can customize as their product evolves.
or
b) S3 - a generic data storage back end that:
* has higher costs than their own engine (according to Nik).
* probably doesn’t provide the functionality required by their client applications both now and in the future.
* Omnidrive would have limited or no control over S3’s future pricing, support, performance, etc.
I know what way I would be going.
[...] In the third episode of Talkcrunch we spoke to the guys from Amazon Web Services and Amazon S3, their new online storage offering. I talked a lot more towards the end, more as a competitor to S3 rather than the podcast host. Everyone else on the show made an effort to convince me that it would be better for Omnidrive somehow if we used S3 as our backend and focused on the front-end. Michael even went as far as calling S3 ‘revolutionary’ which I just don’t agree with - we have had storage API’s around for years. [...]
Nice show. But even I am going to complain about the intro music. It’s really hard on our ears.
I have one more suggestions too. If possible try to limit each show to around 30-35 minutes.
I’m just now trying out Jungle Disk, a free (beta) app that uses S3 as backing store for a WebDAV service. A few setup hiccups, with prompt response from the author, doing a backup of my address book now, haven’t done a restore yet.