As a CloudFlare Optimized Partner, we are thrilled to offer the CloudFlare Railgun™ technology to all our customers. Railgun is CloudFlare’s latest performance optimization technology that gives you significant improvements in site load times. To activate Railgun, log into your control panel and select Railgun “On”. Read on for additional details and instructions:
Railgun ensures that the connection between our network and the CloudFlare network is as fast as possible. Railgun achieves a 99.6% compression ratio for previously uncacheable web objects by using techniques similar to those used in the compression of high-quality video. The average website can expect a 1.43x performance increase.
When a request is made to a CloudFlare server for a web page that is not in cache CloudFlare makes an HTTP connection to the origin server to request the page. It’s that HTTP connection that Railgun accelerates and secures.
Even highly dynamic websites change slowly
Railgun works by recognizing that uncacheable web pages do not change very rapidly. For example, during an experiment, the CNN.com homepage HTML was captured once, and then again after 5 minutes and then again after one hour. The page sizes were 92,516, five minutes still 92,516 and one hour later 93,727.
CNN sets the caching on this page to 60 seconds. After one minute, it is necessary to download the entire page again. However, looking inside the page, not much has changed. In fact, the change between versions is on the order of 100s of bytes out of almost 100k. Here’s a screenshot of one of the binary differences between the CNN home page at five minute intervals. The yellow bytes have changed, the rest have not.
Experiments at CloudFlare have revealed similar change values across the web. For example, reddit.com changes by about 2.15% over five minutes and 3.16% over an hour. The New York Times home page changes by about 0.6% over five minutes and 3% over an hour. BBC News changes by about 0.4% over five minutes and 2% over an hour.
Although the dynamic web is not cacheable, it is also not changing quickly. That means that from moment to moment there’s only a small change between versions of a page. CloudFlare Railgun uses this fact to achieve very high rates of compression. This is very similar to how video compression looks for changes from frame to frame; Railgun looks for changes on a page from download to download.
Railgun consists of two components: the sender and the listener. The sender is installed at every CloudFlare data center around the world. The listener is a software component that we have installed on our network for customers.
The sender and listener establish a permanent TCP connection that’s secured by TLS. This TCP connection is used for the Railgun protocol. It’s an all binary multiplexing protocol that allows multiple HTTP requests to be run simultaneously and asynchronously across the link. Or in other words, a persistent connection is opened between CloudFlare’s network and ours and eliminates the overhead of TCP.
Need an important item delivered now? Royal Mail Sameday® is our fastest and most flexible collection and delivery service. It’s the perfect way to send urgent goods and offers you a dedicated guaranteed same-day delivery.
Peter Hennessy said that he had “grave reservations” about the broadcasting of the leadership debates, which took place in Britain for the first time before the 2010 general election.
Warning that they meant future leaders would be selected for being “plausible tarts” rather than simply “decent”, he said a culture of “celebritocracy” could blight politics.
Lord Hennessy, a cross-bench peer, added that Clement Attlee, the post-war Labour prime minister, would never be elected today because his “considerable virtues” would be eclipsed by his “charisma and presence of a gerbil”.
“The problem with the debates is that the structure of them, the nature of the celebrity in Britain in parts of the media these days, mean that to shine in leadership debates before a general election, you need the characteristics of a plausible tart,” he said.
“Now, being a plausible tart is about 10 per cent of the requirement of being prime minister — no more. It’s not the most important bit.
“My fear is that we’ll be stuck with these leadership elections and when parties choose leaders in the future, the plausible tart bit will play too powerfully in their choice of leader and therefore rule out the decent but non-tarty people.”
“They’re only happy in their own company.
Speaking of the perils of 24-hour news, Lord Hennessy also criticised the habit of parties to rush into responses and “collapse in tittishness” by posting reactions to everything on Twitter.
“The cumulative power of celebritocracy and soundbitery means that it’s an ever narrower spectrum in which you operate,” he said.
“Clement Attlee wouldn’t get through any selection process now because his answers would be monosyllabic.
“He was a man of considerable virtue but it would be unimaginable today, because he had all the charisma and presence of a gerbil. It’s inconceivable now. If there was such a character around, he or she wouldn’t make it.”
Here’s an unofficial but quite professional screencast showing how it should work:
First I try to connect to the printer network:
Then I try opening the print setup:
Then Print Utility and nothing happens apart from the OSX lit up click lamp on the button.which I tried despite the fact that the open printer utility fails (it should launch a web page hosted by the printer itself).
This is the start of the problem: it never seems to know where it is, or if it does it won’t volunteer this information using the Bonjour protocol by default, but it’s address does appear in tcpdump when connected to its own “device” network:
The official way to find this out is via by printing out a test page. The address assigned is always on the 169 subnet and keeps changing. When I apply the changes to the wireless alone or the wireless and what are the correct network settings together it goes in to a hanging mode where the printout shows it has attached itself to my network but its address is still on the 169 subnet and it’s using its own dhcp.
One last thing to try might be to set my own network to the same as this but as there are at several other devices on 192 I’m reluctant to even try that out.
Wikileaks, the site that has infuriated the US government by releasing thousands of US diplomatic cables, is being hosted by one of the symbols of that country’s internet success – Amazon.
The site came under a “Distributed Denial of Service” (DDOS) attack on Sunday night from an unidentified hacker, forcing it to seek a new location for its computer files. And it found it though Amazon’s “Elastic Cloud Computing” (EC2) service, which enables businesses to hire its servers and store their data there.
DDOS attacks typically force sites off the net unless they have enormous bandwidth at their disposal or highly effective countermeasures. Wikileaks, being small and struggling for funds, is neither.
But EC2 allows companies to pay for their usage as it mounts up, rather than upfront.
While Amazon is American, not all of its servers are hosted there – and it could cause a major incident if the US government were to take action against a company on the basis that it might be hosting material the government finds embarrassing.
Brilliant example of Twitter’s idiot proof design :
Sorry, ignore the highlighted words – I was using that as a lazy way to post this blog using the excellent Press This WordPress bookmarklet, and the tweet itself was lazily created the same way. FWIW I resent the tweet as the first attempt resulted in a fairly dire server busy error. Maybe its that unreliability that inspired the sanity check loop?