Hosting Websites on Bare Minimum VPS/Dedicated Servers

Encept – $6.50 128MB FreeBSD/Xen VPS in Ashburn

Tags: , , , Date/Time: April 28, 2011 @ 9:21 am, by LowEndAdmin

Encept Justin, the founder and admin of Encept wrote in to offer some special promotion for LowEndBox readers. Use this sign up link to purchase the following Xen VPS package for $6.50/month:

  • 128MB memory
  • 10GB storage
  • 50GB/month data transfer
  • 2x IPv4 + /96 IPv6
  • Xen HVM/SolusVM

Regarding to the monthly data transfer, Justin mentioned that they are actually quite soft over the overage, and typically would not charge customer unless into the TB range. They accept credit card only via Google Checkout and no PayPal. Servers are with Latisys in Ashburn VA, although Justin also mentioned that they have nodes in Los Angeles.

While Encept.com itself was pretty new (registered in Jan 2011), it is part of Zeruck LLC and Justin has been providing service at Oss.im for 4+ years. They took pride in themselves on FreeBSD, and offer FreeBSD 8.1 under Xen HVM. They also got quite a few recommendations on WHT. The price is far from being the cheapest, but still a reasonable price for those here who prefer Chuck over Tux :)

53 Comments

  1. Youkai:

    A /96? I have to wonder why they do that. They should trivially have enough address space to give everybody full /64s or larger, even if they only have a /48 themselves.

    April 28, 2011 @ 11:08 am | Reply
    • rds100:

      If they rent a server and only got a /64 with it (due to datacenter policies or whatever) – it would make sense to give only a /96 to customers. I don’t know why anyone needs more than a /128 anyway.

      April 28, 2011 @ 1:35 pm | Reply
      • Additionally, we own all of our hardware and colo it based on where we see the need. Latisys, Ashburn used to the Pryme, known as Pull the Plug. We are visiting two datacenters in Dallas, Texas (tomorrow, Friday, April 29, 2011) to determine if they meet our needs. We also have nodes sitting at SafeHosts in UK, just waiting on upstream providers and some last minute hardware to ship.

        Despite having a /32 of IPv6. Its just easier for /96s. Most don’t use more than 10-20 addresses any way.

        April 28, 2011 @ 1:41 pm | Reply
      • Youkai:

        In that case, the appropriate thing to do is request a block for customer delegation. The datacenter should trivially have enough space to give everybody /48s or more.

        As a general rule in IPv6, if your subnet size isn’t /64, somebody’s stuffed up somewhere.

        April 28, 2011 @ 4:56 pm | Reply
        • rds100:

          Yes, and the data center can say “sure, it will cost you $$$”. And it is hard to make a justification for such an expense.
          IMHO at this stage of IPv6 penetration a /128 would be enough for 99% of the users.

          April 28, 2011 @ 5:33 pm | Reply
        • SwordfishBE:

          Little freedom! /124 , 16 ips to get started.
          /96 (4.294.967.296 ips) is more then enough.

          April 28, 2011 @ 9:06 pm | Reply
  2. May I know what is the good thing about FreeBSD?

    April 28, 2011 @ 1:12 pm | Reply
    • Ahh, the good age-old-question about the good thing about FreeBSD, thats really a matter of opinion. I’d suggest going out to the Google monster for that one. But in terms of a test ip, etc:

      IP: 69.42.208.12 | 100mb.bin | 1gb.bin

      Also, the FreeBSD VPS are hosted in LA with AIIX.

      April 28, 2011 @ 1:34 pm | Reply
    • circus:

      Ports and FreeBSD Handbook.

      April 28, 2011 @ 3:31 pm | Reply
    • The *BSDs are more of a direct descendant on the original UNIX while the Linux group of OSes are more of a fifth cousin twice removed that were written from the ground up without any of the original UNIX code. (OK, that’s actually iffy since UNIX code has been discovered in there but you get the idea.)

      It’s supposed to be more stable as well as more secure. Opinion differs on that though.

      April 28, 2011 @ 3:46 pm | Reply
      • Daniel @ LowEndTalk:

        Also, BSD is the basis for Apples Open Source Kernel (Darwin)

        April 28, 2011 @ 5:37 pm | Reply
    • A better question is, “What is the bad thing about FreeBSD?” :) Sorry I have conformational bias towards BSD, even coming from a SCO/AIX administrator background. I hate AIX :)

      April 28, 2011 @ 6:30 pm | Reply
      • I hate SCO :) Worked as a developer/admin guy for some SCO OpenServer/UnixWare boxes back in the late ’90s.

        April 29, 2011 @ 2:02 am | Reply
        • Darren Young:

          I agree. I was a AIX developer and HPUX administrator for a decade or so. /me killed himself every night, but for some reason I woke up every morning… hated it.

          April 29, 2011 @ 2:47 am | Reply
        • Ditto but it;s what I came up with. I feel more comfortable with the BSDs. I know the commands more so that with Linux based OSes.

          When ever I;m going server work, I’m flipping back between putty and google and my notes.

          April 29, 2011 @ 9:29 pm | Reply
    • No-one has said it so I think someone needs to make the comment that FreeBSD is definitely better. BSD is a better breed. I suggest you start using it and find out for yourself. You may find some things harder but overall I expect you will find most things are also better. There’s no substitute for quality.

      April 30, 2013 @ 6:58 am | Reply
  3. Also, is there any test IP?

    April 28, 2011 @ 1:12 pm | Reply
  4. @Youkai: To solve your wonder, we issue a /96 to each guest operating system because we only issue a /64 to each host’s VLAN. Its much easier, efficient and less confusing to most* of our end users to give them a block size /96 (although I admit a /64 would be nicer for each user), versus giving them (random) IPv6 IPs, Forcing them to IPv6_4 tunnel hike it, etc.

    Hopefully this helps solve the mystery behind our veraciousness IPv6 allocations ;)

    April 28, 2011 @ 1:31 pm | Reply
    • Youkai:

      Right, I see how you got there.

      I’m pretty sure the right thing to do is to continue using the /64 on each VLAN, but then route a separate, larger block over it. (e.g. Assign your upstream router to ::1/64 and the VM host to ::2/64, and then route a /48 to ::2). Then you put a /64 on each HostVM virtual interface, with the same ::1/64 = Host and ::2/64 = VM setup, and then route a /56 (out of the /48) to the ::2.

      That way you have nice /64s on each link, but each of your clients gets their own /48 to do whatever with — say, route down a VPN, or whatever. No tunnels or random (SLAAC?) IPs needed.

      April 29, 2011 @ 3:21 am | Reply
  5. Tom Kenney:

    Ordered a 512MB box from Encept back when they were called Host Ossim in February of 2009. Only had two down times, both planned and then a time when I screwed my own box up. Going to grab up another with this promo, its $3.50 cheaper than they’ve ever had it on their site!!!!

    April 28, 2011 @ 1:54 pm | Reply
  6. BSDmonkey:

    Been around the block with freebsd and openbsd vps providers, a lot have come and gone. Signed up about two hours ago via LEB and paid. Provisioning was not instant, which was fine, took all of fifteen minutes. Installed and initial configuration done. Very fast on the disk i/o and cpu time. Network seems stable and is only a hop from one of the freebsd ftp mirrors!

    Will update in a few months once I really get a feel for it. So far, Thumbs up!

    April 28, 2011 @ 5:27 pm | Reply
  7. MBGear:

    Could someone update me on FreeBSD’s performance under XEN. I had heard ( never verified it ) that FreeBSD – XEN was not the perfect combination. Maybe version 8 has resolved these issues ( don’t know ). To note that they offer version 8.1 ( latest is 8.2, released this year ).

    April 28, 2011 @ 5:30 pm | Reply
    • I feel the performance with FreeBSD under Xen-HVM is very good. If it wasn’t, we wouldn’t be selling it :)

      April 29, 2011 @ 2:19 am | Reply
      • Darren Young:

        Hey Shat/Justin:

        I was referred to this post from another friend who apparently already has your service. Going to be signing up soon. Just waiting to get some finances in order and cancel another VPS I hate.

        April 29, 2011 @ 2:49 am | Reply
  8. I have replied to most all the comments and questions related to the Encept promotion, just awaiting moderation because I used my Encept email address initially. But here is the test ip and files:

    IP: 69.42.208.12 | http://69.42.208.12/100mb.bin | http://69.42.208.12/1gb.bin

    The notes on the /96 versus /64 etc are following, once other comments are moderated by LEB admin. We’ve thought about changing over to /64 allocations, but either way, its native IPv6 and no one has complained that is currently consuming our resources.

    April 28, 2011 @ 5:31 pm | Reply
  9. BluePanda:

    How many IP’s is /96 IPv6?

    April 28, 2011 @ 5:41 pm | Reply
    • Hmmm, it must be 2^32 if I am not wrong

      4294967296

      April 28, 2011 @ 5:52 pm | Reply
      • SwordfishBE:

        That’s correct :)

        April 28, 2011 @ 9:07 pm | Reply
  10. @BluePanda: too many to count.. lets just say more than you’d ever need :)

    April 28, 2011 @ 5:48 pm | Reply
  11. Going to be some duplicate comments and I apologize to LEB admins in advance. Trying to answer these questions and I have, but awaiting moderation on most comments and finally figured out which name/email/web combo I used in the past for comments. The test ip is: 69.42.208.12

    We use /96 because we allocate a /64 to each host node. Our upstream gives us a /32 but we still aren’t perfect at IPv6, we’ve thought about migrating to /64s but no one has complained thus far. If you’d like a /64 instead of a /96 we could accommodate you, just have to put it on a different vlan :)

    April 28, 2011 @ 5:52 pm | Reply
  12. Greg:

    @Shat: what would be the addtl cost to double the RAM to 256 and leave everything else in the package unchanged? Also, are nodes available in Los Angeles? TIA

    April 28, 2011 @ 6:06 pm | Reply
    • Greg, all of these nodes are in LA. We’re not putting customers on the Ashburn/Latisys network right now (sold out). However, there was a miscommunication between Admin of LEB and us. Send us an email to info@encept.com and I’ll calculate the deal with you.

      April 28, 2011 @ 6:26 pm | Reply
  13. BluePanda:

    So let me get this right, if i buy this plan i get 4294967296 Ipv6 IPs..???

    April 28, 2011 @ 6:07 pm | Reply
    • rds100:

      And then you run out of RAM and everything else if you try to use even 0.1% of that 4 billion ipv6 addresses :)

      April 28, 2011 @ 6:11 pm | Reply
    • Ha panda. Most customers utilize ~ 10-20 ipv6 addresses. @rds100 is right. You’d run out of ram if you tried to use too many. You can also speed test from http://67.220.77.10/testfile which is a customer of ours who just emailed said he put up a test file on his Archlinux VPS we host. 6.90MB/s from cachefly.

      April 28, 2011 @ 6:27 pm | Reply
      • BluePanda:

        Even if you just do a bind all nginx server?

        Sorry but im very new to actual networking. Its the actual software that i work and help people with so bare with me..

        What is the point of offering so many IP’s if the VPS can only use a few?

        April 28, 2011 @ 7:15 pm | Reply
        • Justin Capella:

          Better than the alternative of having fewer addresses than you can use ;) as was the case with IPv4. IPv4 addresses use 4 bytes, IPv6 use 16, which means there are 2^(16*8) addresses. When they ran out of IPv4 space, they didn’t want it happening again, so now there is an over-abundance. And in theory the VPS could easily use every one of the addresses, but why anyone would need to is somewhat a mystery. If you’re interested in all that stuff maybe check out http://en.wikipedia.org/wiki/IPv4_address_exhaustion

          April 28, 2011 @ 7:25 pm | Reply
        • rds100:

          Each ipv6 has to be “programmed” in the VPS kernel. Each ipv6 address is 128 bits (16 bytes). The normal implementation would be for each interface to have a pointer to a linked list of addresses, where each entry would contain address type, address, mask, pointer to next in list and other info – so atleast 32 bytes for this structure, probably more.
          Now 32 bytes x 4 billion addresses = 128GB of RAM just for these data structures. Now imagine walking all this list of 4 billion entries…

          April 28, 2011 @ 7:27 pm | Reply
      • Justin Capella:

        (That would be me =p). I just re-ran the test, with -O /dev/null. And got even faster (10.1 MB/s) or ~ 80.8mbit/s. And thats while running other services mind you.

        I’ve been with Encept for a couple months now, and they are AWESOME. They over deliver in ever aspect, service/support, CPU performance (Justin even threw in an additional core, because these boxes are UNDERSOLD!). It should be noted that its not exclusively FreeBSD, there is a wide choice of flavors of both bsd/nix. I run Arch on mine. And the images are all up to date. I’ve been with several low-cost VPS providers, and they were all garbage, images from 2005, kernels from the same, LITERALLY DAYS to compile anything. Support staff that might get back to you in a week, if you were lucky. Definately not the case with Encept, I’ve left my number in a support ticket, and recieved a phone call minutes later, it was Justin, informing me he had taken care of it.

        The control panel, dubbed “EnControl” makes sure you are just that – in control – has all the latest features, including VNC access – which is sweet if you have a wtf moment configuring sshd/keyfile. I don’t know what else to say really, other than you’d be foolish not to take them up on this offer. I’m urging several clients/friends of mine to ‘make the switch’.

        April 28, 2011 @ 7:18 pm | Reply
        • ab:

          First host I’ve seen to actually skin solusvm, but try port 5656 and have proper SSL :)

          April 29, 2011 @ 2:24 am | Reply
  14. Just added FreeBSD 8.2 release to installable images. Additionally, all of our isos are netinstall/bootonly. This insures you get the most up to date packages during installation. Just a heads up.

    April 28, 2011 @ 6:38 pm | Reply
    • Daniel @ LowEndTalk:

      If you could add PureDarwin, that would be awesome :)

      April 28, 2011 @ 6:41 pm | Reply
  15. Whoo! I have come around leb for quite some time and never thought I would see the OSSIM guys on here. Crazy. I have been with OSSIM, now ENCEPT as of 2011 for about nine months. Its been freaking ossim! I miss the old name. I have 512MB VPS in LA and a 2GB VPS in Virginia on the other network. Both are blazing fast!!!

    All I have on this subject.

    April 29, 2011 @ 2:45 am | Reply
  16. faizal:

    do you accept paypal? or could gco using paypal? never use gco before..

    April 30, 2011 @ 3:45 pm | Reply
    • No we don’t accept paypal. GCO is just our merchant provider. If you have a paypal debit /mastercard you could use it. Sometimes we make special arrangements for clients needing to pay with paypal but it is rare because it’s a billing nightmare.

      April 30, 2011 @ 3:57 pm | Reply
  17. Last weekend we racked up new nodes in Dallas, Texas at CoreXchange. Anyone interested in an early bird special? Let us know via email info@encept.com letting us know, it’s for today and tomorrow only.

    256mb memory
    20gb disk
    1 ipv4
    200gb bandwidth

    5.50/month.

    May 6, 2011 @ 3:27 am | Reply
    • Ian L:

      Been curious to test out CX’s b\w mix and such. E-mail is on its way…

      May 8, 2011 @ 1:59 am | Reply
      • Correction, it is still $6.50/monthly, not $5.50. The early bird is to get the double ram and extra disk and bandwidth. If you’d like to order this deal you can email info@encept.com and we will let you know the availability.

        May 10, 2011 @ 6:14 pm | Reply

Leave a Reply

Some notes on commenting on LowEndBox:

  • Do not use LowEndBox for support issues. Go to your hosting provider and issue a ticket there. Coming here saying "my VPS is down, what do I do?!" will only have your comments removed.
  • Akismet is used for spam detection. Quoting webhostingtalk.com URL seems to get binned consistently here, but I do peek into the spam box frequently to publish those comments.
  • Use <pre>...</pre> to quote the output from your terminal/console, or consider using a pastebin service.

Your email address will not be published. Required fields are marked *