Engine Yard Cuts 15% Of Workforce | Hacker News Reader

Engine Yard Cuts 15% Of Workforce

techcrunch.com

26 pointsqhoxie18 comments

Yes this was mostly support staff. The main reason for this is that we have become much more efficient and automated so we did not need as many support staff anymore to support our systems.

Well, this certainly flies in the face of the sensationalist tone of the linked article.

TechCrunch Layoff Tracker? Seriously?

"confirm that 12 out of 82 people have been let go"

They had 82 people??? Wow. That seems a lot for a relatively young startup.

I could be wrong but I imagine support staff accounts for a large portion of that.

They have a lot of support staff (they respond almost instantly to problems) and also do a lot of Ruby development and make a lot of open source contributions, etc.

Often, I wonder if this kind of public news hurts them (or any company) overall.

Would you be less tempted to use Engine Yard for hosting knowing that they just laid off some folks? Even slightly?

And if so, would it be worth keeping these folks around just so you aren't making "negative press" for yourself?

I think it all depends on wording. EY's official blog statement talked about helping the employees find new jobs, which is impressive. To me, it would also be beneficial for them to include a statement like Ezra's above, to ease customer or potential customer concerns: They laid of support staff because they have further automated their environment - that is reassuring.

As a customer I'm glad they are laying off people. They charge far more than they should for the resources you get, and it seemed like they had way too many employees than they needed. Our sales rep was totally clueless and gave us bad information several times - I wonder if he is still there..

Anecdotally, I'm shopping around for a Rails host right now and being reminded of Engine Yard because of this article increases my chance of going with them.

If the layoffs had been 50% I would be worried, but at this level it doesn't trouble me as a potential customer.

Recently moved a client off engine yard that was paying ~$2600/month they are now having substantial savings for more capacity with EC2/AWS.

We're paying around $1100/month and looking to migrate off EngineYard as well - can you point me in the right direction and/or provide any tips for a migration to AWS?

Tips:

* Automate Server Configuration with Puppet

* Store your database on an EBS

* Backup often to S3

* Use cloudfront to host your most common static files (CSS, JPG, JS)

* If you don't have too much email volume, use google apps for your domain.

* Use DNSMadeEasy.com for your DNS and also failover to your backup.

* Anything that you didn't automated, document it so that its easier to automate later.

* Have a backup plan if EC2 goes down.

If you want some help with this, you could always email me dan {at} abtain {dot} com.

How sad is it when EC2/AWS represents a substantial savings over your service?

Why is it sad? Don't forget that using EC2/AWS isn't free either. There are costs to migrate to that architecture, to maintain it, to debug weird bugs that will inevitably arise, etc.

Those problems also exist on an infrastructure like EY's, but part of the price you pay to EY is to have a team of people who deal with those problems for you.

I'd be interested in knowing whether quellhorst's fees are included in the "substantial saving".

Yes, my fees are included in the savings. After the setup fee I'm charging $300/month on top of the AWS fees to keep things running. If the client needs things like performance optimization, that would be billed hourly.

Its sad because Amazon itself is a significant premium to rolling your own host for most situations. (It's undoubtedly worth it often, but in terms of hardware cost, it's pricey.) So EY is a significant premium to a significant premium.

It's not that pricey. Let's Price EC2 against SoftLayer (which are actually the two most common hosts for YC companies).

SoftLayer 4x2GHz cores 16GB RAM 2TB Bandwidth 250GB RAID 1 storage 500GB backup $999

EC2 4x2.4GHz cores ($576) 15GB RAM 2TB Bandwidth ($340) 250GB EBS storage ($25) Backup (~$50) $991

To be fair, they aren't equivalent systems completely.

* Amazon you can scale up or down your bandwidth and just pay for usage.

* SoftLayer you have a 1,000Mbit port rather than the 250Mbit port in EC2.

* Amazon you can scale storage effortlessly, SoftLayer you need physical drives or to pay for their NAS at $0.40/GB.

* Amazon you can provision in minutes, SoftLayer takes hours.

* Amazon snapshots are easy, don't know about SoftLayer.

* SoftLayer often runs promotions like double RAM and double HD so you can knock the price down by $120/mo based on that.

AWS isn't cheap, but it's easy. SoftLayer is offering a great deal too, but I like (at least the concept of) being able to launch a new box within 5 minutes if one dies and being able to just say, "hey, do a delta snapshot to S3, thanks!" rather than dealing with creating a reliable backup strategy. I do love the idea of getting a 32GB RAM server for the price of a 16GB one and a 1,000Mbit port, but I don't want to give up the ease that Amazon offers.

So, AWS isn't so out of the ballpark. There's a premium for the complete flexibility to scale storage from 1GB to several TB (or even petabytes with S3) and the ability to launch instances on demand and backup with no hassle and I'm willing to pay that if it's reasonable and AWS is reasonable.

I don't know where your Softlayer pricing comes from. We have two 2.33ghz 8 core servers with NAS for just over what you quoted for one 4 core. We're db-heavy though, and not much in the way of file serving so we're lighter on storage (we have 2 73ghz 15kRPM SCSI disks on RAID 1) but that doesn't cost much.

Most startups don't do anything that requires them to significantly scale storage, and if they do the transfer costs on Amazon are huge (though quite possibly still worth it). If you're running Dropbox or Flickr type stuff, Amazon's amazing.

AWS would be an order of magnitude more for startups like us that are basically MySQL-based. There are a lot of us out there.

Interestingly, startups that go the traditional route are limited in that they have to have enough server capacity to meet peak demand on hand 24/7, so for much of the day you're paying for idle servers. I imagine Amazon becomes a lot more price-efficient if you have a high peak to trough traffic ratio.

Reply on news.ycombinator.com