Startups: Don't Host Your Blog on Tumblr
blog.picplum.com
blog.picplum.com
In 2011, Tumblr was down for 42 hours.
In 2011, Heroku was down for 76 hours in one incident alone.
And that's all before you get to the hubris of the assumption that your custom coded setup will perform better. A more sensible argument would be to move to a host with a better proven uptime record, such as Blogger (nearly 100%).
Sources: http://allthingsd.com/20111215/tumblr-had-42-hours-of-downti... https://status.heroku.com/incident/151
- I wanted to use Sinatra to 301 old tumblr urls
- Call me weird but I loathe seeing *.html in urls and that's the only thing you can do with S3 root object (when I checked a few months ago)
- something I can't remember. maybe it was cf edge server time to update when altering updates.
You can configure name of index file, and that will be displayed for requests against root URL and subdirectories (such as blog posts) as well. No need for having .html in URLs.
The only problem I've encountered with S3 web hosting is that it does not let you host naked domains.
OK that was actually the answer.
Honestly there's a little bit of "because I did it myself and I'm awesome it won't fail." That's hubris and very typical of young engineers.
Letting someone else manage it for you is probably the right thing for a startup to do. You just need to choose someone with a good, proven uptime record. That is what should guide your decision.. with so many things in engineering.
I'm not saying you don't have a point, but imo your comparison of Heroku's 76 vs. Tumblr's 42 is not apples-to-apples. I think you have to at least put an asterisk by the 76. What OP is complaining about is a early-Twitter-days-like unreliability over a long interval, not a one-time outage that has a single explanation and for which Heroku took completely responsibility.
For my part, at the moment at least, I still have more confidence in Heroku than Tumblr, 76 vs 42 notwithstanding.
My point is if uptime is your goal you should go with a proven solution. Relying on an unproven custom setup to be 100% mistake free smacks a bit of young engineer hubris.
tumblr has a really amazing community. tumblr's downtime is a fact. The question is, does the community outweigh the downtime?
I think you make this decision based on the demographics of the tumblr community vis-a-vis your users / customers. And tumblr isn't just 14 year old girls -- there's a bunch of tech people on it (although that may be more NYC tech).
In the comments, people have mentioned the simple fact that you can have both a self-hosted company blog and a company tumblr, much like most companies have a company blog and a company twitter. I agree, although simply feeding content from the blog into the tumblr of course doesn't make best use of tumblr as a platform.
My startup's blog is on tumblr (http://blog.idonethis.com) and we get about 1/5 of the referral traffic of twitter and facebook from tumblr. (We have nearly 1k followers on tumblr versus 1.6k twitter followers and 2k facebook likes.) In other words, it's our 3rd largest source of referral traffic from social media.
Tumblr is comparable to Twitter in that it's a common place to meet your readers/customers. Unlike a self-hosted blog, you don't have to ask visitors to leave where they are to visit you. So you're likely to get a bit more foot traffic/interaction, not to mention people passing it on/sharing it out (via reblogs).
Now if you're an already established service/otherwise famous, you might have no problem getting people to check you out. But even then, Tumblr has its benefits.
I don't think the choice is necessarily all or nothing between self-hosted or Tumblr. Many companies with "real" blogs also have Tumblrs. In many cases companies have multiple Tumblrs. (Not unlike Twitter, it's becoming common to have a separate Tumblr for different events/products/branches/divisions. IBM has at least 3. NBC has at least 4...)
I won't disagree that Tumblr has horrible performance issues. But it's still worth it. Twitter 3 years ago had bad performance issues, but it was/is still definitely investing in. No social site is forever, but they're great tools as long as you're aware of costs + benefits.
From their about page:
"What is Tumblr? Millions of people sharing the things they do, find, love, think, or create."
Even if you can do more than a simple blog with tumblr, at its base it is a blogging platform. It's perfect for a startup that is busy doing lots of other things... if it wasn't for the fact that it's always down. I hope they'll someday add a "Always Up and running, no trouble" to their "30 reasons you'll love tumblr".
Since most startups would be using a subdomain of their own domain for their tumblr blog, so I don't think this is an issue.
As much as this is a very staunch gesture, I think it's a very valid one. Tumblr has a huge, active, and connected community. It can be a great entry point for niche startups to find users and build an audience, but the performance and stability just does not cut the mustard.
I greatly want tumblr to succeed, their platform is very fresh, the theme garden model they have fits their mold perfectly (although the premium marketplace seems to be extremely esoteric) but I also hope this suggestion for startups to choose an alternative platform kicks tumblr in some sort of direction to improve the QoS.
Not perfect, of course, but it still gives you a slight chance to use Tumblr's amazing sharing platform.
This kind of reinforces my earlier commentary, tumblr has some great features and awesome integration, but their performance and reliability always make me steer clear when needing to put up a blog for a client.
For most startups the amount of time invested for a custom Jekyll setup and a downtime script could be replaced with a new feature or a better product/home page.
I think its important for startups not to spend too much time blogging unless they feel it will provide them some decent traffic and strongly support the product/company. Make the product better first. If your blog host (Tumblr) isn't stable enough, wordpress it and be done with it.
'more features' != 'more marketshare'
See 37signals blog, seomoz blog, and every piece of advice patio11 has ever written.
"unless they feel it will provide them some decent traffic and strongly support the product/company"
Not sure you added anything to my previous comment.
It's like having an @aol.com email address rather than your own domain. It looks unprofessional at best.
And in the case that downtime is as bad as it seems, I suggest you guys use CloudFlare, which will at least keep a cached version of your site online even during downtime issues.
Tumblr is way more about community than mission-critical uptime. If you treat it like mission-critical uptime, you're going to be disappointed. If you're trying to reach an audience for your application — which, honestly, should be more important with a staff blog — Tumblr is a great bet.
The only solution that I know of is to host it all on the one box which makes it trick if you want to use Heroku to host your app.
I am currently wrestling with this problem as I want to host a new app on Heroku, but have the marketing side of the site powered by wordpress to take advantage of things like third party themes. The app will also generate linkable content, so it is important to me to have it on the same domain, but there seems to be no way to achieve what I want to do with heroku :-(
So annoying!
It was actually really easy with this: https://github.com/mhoofman/wordpress-heroku
But unfortuently, to the best of my knowledge, you can't host both PHP at the same time as Rails.
If anyone know otherwise I would love to hear!
I just read the DotCloud docs and they do a lot of stuff including php and ruby, you might want to check them out to see if they run both in the same instance.
I'll ckeckout DotCloud as well. Thanks for the heads up!
edit: added request for more info
If you have to worry about scaling or downtime if a blog post hits HN's front page, then I'm afraid that you screwed up really bad somewhere.
Any server should be able to handle that traffic.
PicPlum's recent blog posts seem to be pretty run-of-the mill ... press mentions, tips, new features, etc. Nothing that (a) would hurt the company if the blog was down or (b) you couldn't simply use your company Facebook page for.
We are big believers in making quality content at Parse as well. Interruption marketing is annoying and not nearly as effective.
Given that, it's pretty crucial for our blog to be up as well. (we're also on jekyll, hosted on our own servers)
Aside from unexpected incidents (and limited offers), you'll be fine. But if something goes wrong, you don't want panic and misinformation to ensue.
It all depends on your kind of company, of course.
Sites (in my experience!) with a lot of these links have way more loading problems. Simple sites that rely more on markup and CSS3 than images and a lot of JavaScript have only dropped out a few times in my experience.
Just my two cents, maybe somebody can comment and verify this.
It's your writing, if it is good and insightful, write it on a server you have control over.
My livejournal blog however, which I set up over 10 years ago (and haven't updated in 7 years) is still there in it's embarassing completeness.
I wonder what accounts for our different experiences?
* Don't blog on Google+
* Don't blog on Tumblr
it doesn't matter how you do it right now, just do it and worry about that later.
A lot of the decision is the 'quality' of the downtime. For example:
* Is it generally stable, but only going down in peak load situations (obviously no good if you know peak load occurs when you are doing critical new product release/sales etc, but may be okay if you peak load corresponds to non-critical events).
* How long is the outages? If it goes down for 1 minute each day, that may be more acceptable compared to '100%' uptime except one day when it goes out for 365 minutes.
All of that affects the impact that the downtime has on your venture - it's quite possible you might have 'poor' uptime stats, but it has zero effect on your business, in which case the cost/effort savings from using a 'poorer uptime' hosted service may be worth it.
Priorities...
If the site is down from 12am to 6am due to some botched maintenance, I really don't care, as traffic will be minuscule, if not non-existent for our core market. And 6 hours of downtime is a killer for uptime percentage.
The "Jekyll is so easy! It's the easiest thing I've ever maintained!" argument reminds me of Andy Rubin's now infamous quote about the openness of Android: "the definition of open: "mkdir android ; cd android ; repo init -u it://android.git.kernel.org/platform/manifest.git ; repo sync ; make"
The point of Tumblr is I don't need to maintain _anything_!
It's a static site, so you can host incredibly easily on any number of cloud services, from heroku to AWS S3/CloudFront, many (most?) people keep their Jekyll sites in a git repo (often github) so you won't lose data, and can easily shove it onto another server in no time at all.
Really, it's less maintenance than pretty much anything I've ever had online.
Costco.com will print for 12 cents per print and free shipping, are you guys going to be able to be competitive against them?
Happy to report that price has not been an issue and we have many happy return customers. Try it out and let us know what you think! Each new account has $5 in credits.
Heroku-hosted custom Jekyll setup
But the whole point of Jekyll is that it generates a static site.
What am I missing? (seriously)
I thought a starter Heroku account might work well for a side project I have cooking... Coincidentally, I was reading about Jekyll as something fun to try. Combining both of them would be pretty cool.
What I dig about Tumblr is how fast Google indexes posts from there due to it's popularity.
Thus, if your looking to HELP get your start-up's name listed in Google fairly quikcly, then a Tumblr or Posterous would be best to use. I posted a Tumblr "About us," yesterday and Google indexed it eight hours later.