New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Linode opens DC in Tokyo
http://blog.linode.com/2011/09/19/linode-cloud-asia-pacific/
Perhaps not LEB-grade prices, but from what I see for Asia and general and for Japan especially, this is a superb deal. Also - first the MountSpot offer, now this - maybe something is changing to the better with DC and bandwidth prices in that part of the world.
Comments
O-M-F-G
I am about to try one xD....
Sounds like a good alternative to those providing services to the Far East Asia. I wonder what the latency would be like for those from Asia Pacific.
Would that also reduce the demand for US West Coast servers? Not until there's a low end provider there I guess
Linode is still only providing 200G bandwidth, that's not a whole lot even by LEB standards
Francisco
I'm surprised an average user would even need that much. 100GB used to be "alot" for a VPS a few years ago.
Last month I used ~150gb on my main VPS, but I used it a VPN to it almost full time. This month I'm around 80gb without any VPN use. (TBH I don't know why it's so high)
Ping from Vietnam is around 180ms, not too much faster than West Coast based providers. (average 200ms, Quadranet is around 19xms)
Ping is very inconstant from Western Europe.
One thing would worry me if I had a server there, Earthquake.
Ha. My BuyVM VPS has 1TB/month data, but just then it says I've used 12TB this month! 200GB/month from Linode would definitely not be enough. </sarcasm>
That's indeed a concern. However with earthquake in Tokyo -- there will be a lot more things to worry about than my damn $20/month VPS. On the other hand, what's the likelihood of earthquake on the other side of Pacific ring of fire, affecting all those LA, Fremont and San Jose data centers?
Well, Japans tectonic activity is probably much higher, it has 3 fault lines going right through it,
I do remember doing a case study on a LA & Kobe Earthquake, I believe that from earthquake alone then it wouldn't effect the DCs, but a tsunami would.
Yeah same happenned to me. Anthony said me that they are doing some experiments :P
Not quite expirements, more I was upgrading the backend accounting code and franned it real good :P With the amount of VM's we have, BW accounting is the most brutal part of the whole system. The original Solus version (that stallion was emulating) was doing 8 SQL updates per VPS IP during updates.
The new edition is doing 1 update per IP as well as a lot of other merged counters to cut down on queries.
I'm sad that I had to reset everyones bandwidth but I'm happy that it's all running like it should and running perfectly now, including ipv6 BW accounting.
Francisco
@Francisco Any reason why I have 500TB bandwidth?
Ping from Sydney, Australia:
Traceroute:
I wouldn't worry about earthquakes in Japan affecting the DC as a whole - yes they have a lot of earthquakes there but they are the most prepared nation when it comes to earthquakes. I wonder however do earthquakes affect disk failure rates in disk arrays?
my own BuyVM/SanJose
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss), Approximate round trip times in milli-seconds: Minimum = 214ms, Maximum = 226ms, Average = 221msLinode Tokyo
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss), Approximate round trip times in milli-seconds: Minimum = 122ms, Maximum = 142ms, Average = 132ms.... Worth a try!
64 bytes from 106.187.33.12: icmp_seq=5 ttl=47 time=594.233 ms
64 bytes from 106.187.33.12: icmp_seq=6 ttl=47 time=513.164 ms
That looks like a routing issue...looks like the route goes over to New York--> California --> Tokyo rather through than a physically shorter route through Eastern Europe --> Asia --> Tokyo
@TigersWay @kiloserve
your posted pings are utterly useless if you don't mention where you are pinging FROM!
Superb deal for the entire Asia Pacific? I don't think so.
For what I can see, traceroute seems to suggest that it's direct link from Singapore to Japan (via KDDI), however the ping time to tokyo1.linode.com is 1.5x the time to fremont1.linode.com, and the download speed from tokyo1 is half the speed compared to fremont1.
Solid ping to Cali
Being from Singapore, here are my results:
Another traceroute from downunder
That traceroute from Optus looks horrible!
@Boltersdriveer
I can't seem to get <100ms with Singnet, however I'm able to get that if I were to do it directly from M1 DC, but that defeats the purpose IMHO.
Please check the post. Those weren't my pings, it was from Daniel in Western Europe.
You skipped your breakfast?
If you have no idea where I was pinging from, then you never read any of my post... But it does not really matters as my point was to show some could easily get a benefit.
@TigersWay when posting useful technical information never make the crucial bit of it rely on people knowing you, or knowing where are you from via your other posts, etc
What about someone who e.g. arrives here from googling something.
@maxexcloo @Zigga -- indeed Optus route is totally crap. I found going through TPG would get pretty good ping (~120ms) whereas going through MCI/Verizon is not as good (~180ms). Didn't know Optus/Singtel is this bad. Looks like it's going to US West Coast, then Hong Kong and then Tokyo...
This sounded exciting... Until I checked the routing.
That's 190ms from the Philippines to Japan. It's only 170ms to their Fremont server.
going through the au isps looks like tpg has the best connectivity. even node is doing 132ms(from melb) and they have a pop in Japan!
@david reverse traceroute could contain the answer, perhaps it goes via USA on the path back to you. But to obtain such a trace you will need to run it on a machine on the other end, and target your own IP. Usually scripts called LG (looking glass) are installed for such purpose by hosters and DCs, but I dunno if Linode has one in Japan or at all.
6MB/sec to the UK (RapidSwitch) and only 10 hops.
4.5MB/sec to .SE
Considerably worse (think tens or hundreds of KB/sec) to .nl and .de.