Thanks to everyone who beta-tested the new LowEndTalk.com over the last week. It was hosted on a $20/month Linode with 512MB of memory, which some people reckon it is not low end enough! Also thanks to Brandon from eNetSouth Cloud who is sponsoring me a new VMWare-based VPS, which I have migrated LowEndTalk to and set it live. Here is the spec of the VPS:
- 256MB memory/712MB swap
- 20GB storage
- 1,500GB/month data transfer
- 1x Intel Xeon X3430 core at 2.40GHz
- Ubuntu 10.04 x86 on VMWare
- San Jose data center
This VPS has special price of $7.99/month on eNSCloud’s website, although eNSCloud has been giving LowEndBox readers promotions such as $6.95/month for 400MB VM or $18.75/3 months for 512MB VM. So far so good — 256MB with 1x Xeon core runs LowEndTalk fine with Nginx proxy to 2x gunicorn/Django backend running OSQA. IO performance is pretty decent. San Jose location would be good for Asia Pacific visitors, and hopefully still usable for our European visitors.
Related Posts:
- 5 Reasons Why You Want a Low End Box - May 26, 2021
- Dead Pool January 2012 - February 2, 2012
- exit(0); - January 19, 2012
Cool. eNSCloud is a great match for LowEndTalk :)
Nice job Brandon. You run a smooth operation, definitely one of my favourite providers. :)
So far I’ve been quite impressed with the eNSCloud VPS I picked up from their last offer. Setup took a few days, but I know Brandon and the eNS team have been busy. The control panel needs work, but the VPS itself is great.
Have two VPSs with eNSCloud. Pushing a project with one and going live within next two weeks. They are great. Brandon is a great guy!
Hi,
Don’t know much about VMWare. Can I enabled Tun/Tap (is it enabled by default)?
thanks
JP
It’s a dedicated-server-like system so you can enable tun support without asking the vendor.
So does PPP dev?and IPsec could be used?
Congratulation LEA!
Maybe you can make an archive of all the old LET content.
They are in http://old.lowendtalk.com/ at the moment. If you come from search engine hitting the old content pages, it would redirect you there.
Very nice LEA, I have tried it, and I got redirected correctly, how did you do it? Are you redirecting all requests to “topic” directory?
Do they still need paypal subscription?
You know you can pay and cancel, right?
My pp account (verified) isn’t associated with credit card.. so no.
What is that 712M swap all about?
No idea. That’s just the way Brandon set up the box. I don’t plan to use that much though :)
My box there shows 499MB RAM and 894MB swap. It works great so I’m not complaining. :)
congrat lea with new box.
Gz you deserve it.
“San Jose location would be good for Asia”…
I’m in China, using the only ADSL provider in Beijing (China Unicom). — It completely doesn’t matter _where_ a box is. It matters _which_ route it takes to get there. Many routes out of China are overloaded on a regular basis.
I have to say “San Jose location would be good for Asia” is true in general for most of the countries in the region.
However as you know China is huge and they only follow their own rules, I don’t feel surprise that most routes are overloaded, or maybe, regulated.
Yes. With 1.3 billion people and a *huge* number of internet users. Say your site just got featured by a local TV station of a regional city of *just* 5 million people. Next thing you know is your data center choking on bandwidth (provided your low end VPS hasn’t been killed yet).
And all the “regulation” bit. Well I won’t talk about it in fear of having LowEndBox.com blocked by the GFW. Or maybe it has already :P
It’s not yes unless you post some anti-CPC articles, LEA. Also I don’t hope that happen in the future.
Actually there’s a 5TB cable between US and China but it’s not enabled to be available to we ordinary citizens yes. Current cables are overloaded and the greatfirewall is a plus in latency – usually 10 – 50ms. :)
LEB isn’t blocked by the GFW yet :)
http://viewdns.info/chinesefirewall/?domain=lowendbox.com
But looks like they are missing, cant contact them and they just dont reply the ticket
Yes, I had a looooong mail chain over weeks to sort out one simple issue :(
Just curious, but why Ubuntu over Debian? You’ve chosen Debian in the past.
Maybe he wants/needs newer packages and debian testing does not get security updates in a timely manner.
I signed up with eNetSouth a couple of weeks ago.
My account was provisioned reasonably quickly (~48 hours), but haven’t had much luck in the way of responses to support requests since then.
However, I can’t fault them on performance or uptime – all good there.
Waiting to see the outcome of the tickets I’ve raised at the moment.
Same here. Hopefully they get the support ticket thing sorted out, I think I have two open right now. The uptime seems good though.
Ditto. Quick responses and provisioning, now there is no communication from there at all. Server is still running fine though! (Albeit my support issues remain unresolved)
I don’t know why, but eNSCloud disk I/O performance is gradually degrading over time. Repeated ticket is not helping anymore. I think the system itself is unable to support the I/O anymore.
I remember about 3 months earlier I used to get some 70 MB/s +
What is your I/O performance stat LEA?
Hi Amyt,
I done a test on my Windows eNS Cloud VPS. Getting an average of 25 MB/s.
This is far better than my test. But still 25 MB/s is quite less for a server – but can work with.
I’ve noticed in the past that disk performance can vary, and at times it can get slow, but overall it’s been ok for me. I used to run hourly disk speed checks, but it’s proved itself well enough that I don’t anymore. I checked just now, though:
/dev/sda1:
Timing buffered disk reads: 100 MB in 1.39 seconds = 72.10 MB/sec
Are you seeing consistently slow speeds, or does it come and go?
This is the speed I used to get earlier. Hourly check is a good idea. Maybe I can also write a cron job that can do the hourly I/O test and put report in a text file. If possible, please share if you have any script like that.
Then if it look bad, I will get in touch with eNS Cloud support again.
You could create a simple script like this and set it to run as an hourly cron job:
Can’t believe it !
Is its time to change the vendor?
—
root@srv3:~# hdparm -t /dev/sda1
/dev/sda1:
Timing buffered disk reads: 2 MB in 6.07 seconds = 337.66 kB/sec
Just did a test on my VPS there running LowEndTalk.
1073741824 bytes (1.1 GB) copied, 10.193 s, 105 MB/s
This is quite good!
No one is answering support request. My ticket (High priority) left open for weeks.
Also seems like they have poor I/O in chicago DC (not in San Jose).
Anyway, I am still waiting considering the holiday season.
Later on I will try to make a post on low end talk about VMware VPS issues. I think it has never been hi-lighted well.
LET down today?
No go for me either.
Does this mean I have to go and actually do support tickets now?
My instance has been up and down like a yo-yo in the last 48 hours (San Jose). Have raised a ticket, but haven’t heard back from support at this stage – anybody else having issues?
My VPS in Chicago was going up and down yesterday. Today it went down once, which lasted 11 minutes.
Heard back from ENS a little while ago – evidently some servers in San Jose failed recently & restoration is currently underway. :-)
I’m going through withdrawal :(
Francisco
I KNOW.. O_O
Its baaaaaaaackkk
I got the email, too. I wonder what hardware failed, since one of the benefits of having a Cloud VPS is supposed to be automatic failover to another node during hardware failure. I would think it would only be down if both the primary and backup node(s) failed, or the SAN, or a router failed.
Just to let everyone know what happened we had a failure with several servers in SJC. It’s hard to fit a memory demand of 280GB in 160GB. We have around 320GB worth of RAM across the hosts at 100% capacity. The up and down was due to HA trying to load everything out but there was not enough resources. A lot of times instances were up but the host was so overloaded it was useless.
We can sustain a loss but not several at a time. We are back at 100% now and I appreciate everyone being patient and understanding. This has not been a #winning (Charlie Sheen pun) 30 days for us. We will turn it around though. :)
Brandon
Yeah it’s good that the box is coming back online. Sh!t happens — even over the weekend! I emailed Brandon when I spotted the VPS went down and got pretty instant replies. So the support is all good.
However the machine did struggle a bit over the next 24 hours though and it went up and down like yoyo (actually more down than up). I haven’t bothered to jump up and down because it was weekend, where I am supposed to play with kids, walk in the park and go to church :) Anyway. Glad that it is all sorted out now (finger crossed).