LowEndBox - Cheap VPS, Hosting and Dedicated Server Deals

Review: CastleGem 256MB KVM (7 Months)

Here is a LowEndReview for you…

CastleGem are a London based Internet Service Provider and a software consulting & engineering company, who have been supplying IT solutions since 1995. They’ve been featured twice on LowEndBox and twice on LowEndTalk.
 
CastleGem provide KVM, XEN and OpenVZ services in the UK, Austria, Czech Republic, Germany and United States. They also offer many other hosting services, such as dedicated servers, web hosting and colocation.

 

Disclaimer:

I am in no way affiliated with CastleGem but I have been a customer of CastleGem since July 2012 (7 Months) and this server is powering several websites of mine. Therefore it should be noted that the below results are of a server currently being used.

Introduction:

As mentioned above, I have been a customer of theirs since July 2012 so I believe this review should paint a pretty accurate picture of their service. Spirit introduced them to me when I was looking for a new VPS – so thank you :) I have two servers with them, one located in Vienna, Austria (Xen PV) and another located in London, UK (KVM). The service I will be reviewing is the London based KVM that I have with them, if you’d like some benchmarks etc of my Austrian server, just ask below in the comments!

In case you were wondering, this is a customised LowEnd KVM plan with 512 MB RAM (even though I ordered 256 MB!) and has 5GB of encrypted LVM storage running Debian 6 32bit. As it’s customised, some of the benchmarks may be lower than their standard plans – on signup I mentioned I didn’t need too much resources.

Support & Communication:

Support and communication is one of the most important factors to me. Now I don’t need need my hand holding but it’s nice to know there is someone at the end of an email, ticket or phone call. When there have been issues, few and far between, CastleGem have used both Twitter and Email to keep their clients informed and updated. I haven’t actually needed to use the support, as everything has ran smoothly and it hasn’t been necessary for me to alert them to an issue. I have however corresponded with Billing a few times. On one such occasion I had forgot to pay an invoice. But they didn’t suspend or terminate my service after I had forgot to pay, in fact they only contacted me four days later with a friendly email informing me the invoice hadn’t been paid! It is always nice to see when a provider treats you like a person – we all make mistakes :)

Hardware Information:

For the several LAMP websites I use this VPS for, I barely use much of the below CPU. I’m on an older node with a e3-1230v1 but their new nodes are running the e3-1230v2.

root@ldex:~/benchmark# cat /proc/cpuinfo 
processor	: 0
vendor_id	: GenuineIntel
cpu family	: 6
model		: 13
model name	: QEMU Virtual CPU version (cpu64-rhel6)
stepping	: 3
cpu MHz		: 2494.340
cache size	: 4096 KB
fdiv_bug	: no
hlt_bug		: no
f00f_bug	: no
coma_bug	: no
fpu		: yes
fpu_exception	: yes
cpuid level	: 4
wp		: yes
flags		: fpu de pse tsc msr pae mce cx8 apic mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm up pni cx16 hypervisor lahf_lm
bogomips	: 4988.68
clflush size	: 64
cache_alignment	: 64
address sizes	: 36 bits physical, 48 bits virtual
power management:

Meminfo Results

root@ldex:~/benchmark# cat /proc/meminfo
MemTotal:         514728 kB
MemFree:           29048 kB
Buffers:           53268 kB
Cached:           377040 kB
SwapCached:          124 kB
Active:           251000 kB
Inactive:         208952 kB
Active(anon):       7376 kB
Inactive(anon):    22416 kB
Active(file):     243624 kB
Inactive(file):   186536 kB
Unevictable:           0 kB
Mlocked:               0 kB
HighTotal:             0 kB
HighFree:              0 kB
LowTotal:         514728 kB
LowFree:           29048 kB
SwapTotal:        258040 kB
SwapFree:         257916 kB
Dirty:                 8 kB
Writeback:             0 kB
AnonPages:         29528 kB
Mapped:            31112 kB
Shmem:               140 kB
Slab:              21796 kB
SReclaimable:      17800 kB
SUnreclaim:         3996 kB
KernelStack:         584 kB
PageTables:          732 kB
NFS_Unstable:          0 kB
Bounce:                0 kB
WritebackTmp:          0 kB
CommitLimit:      515404 kB
Committed_AS:      64692 kB
VmallocTotal:     503812 kB
VmallocUsed:        5832 kB
VmallocChunk:     487176 kB
HardwareCorrupted:     0 kB
HugePages_Total:       0
HugePages_Free:        0
HugePages_Rsvd:        0
HugePages_Surp:        0
Hugepagesize:       4096 kB
DirectMap4k:       12276 kB
DirectMap4M:      512000 kB

Inode Allocation:

root@ldex:~/benchmark# df -i
Filesystem            Inodes   IUsed   IFree IUse% Mounted on
/dev/mapper/ldex-root
                      296000   36442  259558   13% /
tmpfs                  64341       4   64337    1% /lib/init/rw
udev                   63041     524   62517    1% /dev
tmpfs                  64341       1   64340    1% /dev/shm
/dev/vda1             124496     222  124274    1% /boot

vmstat:

root@ldex:~/benchmark# vmstat
procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----
 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa
 0  0    124  29048  53300 377060    0    0     0     3   12    7  0  0 100  0

The Network

Without a doubt CastleGem’s London network must be one of the best networks I’ve used in the UK. Most of my traffic seems to go over level3, including to my home, which means speeds are very fast and pings are much lower than other bandwidth providers. As you can see from below, I am able to use pretty much most of the 100mbit port. The network has been solid with no outages as far as I can remember.

Each test was run three times and the middle ranked test was picked.

The CacheFly download speed test showed great speeds.

root@ldex:~/benchmark# wget -O /dev/null http://cachefly.cachefly.net/100mb.test
--2013-02-16 17:58:02--  http://cachefly.cachefly.net/100mb.test
Resolving cachefly.cachefly.net... 205.234.175.175
Connecting to cachefly.cachefly.net|205.234.175.175|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 104857600 (100M) [application/octet-stream]
Saving to: `/dev/null'

100%[================================================================================>] 104,857,600 16.1M/s   in 6.5s    

2013-02-16 17:58:09 (15.5 MB/s) - `/dev/null' saved [104857600/104857600]

For more accuracy, I ran a few more speedtests which were also good.

Hosted by Vodafone UK (London) [1.47 km]:
Testing download speed........................................
Download: 72.77 Mbit/s
Testing upload speed..................................................
Upload: 35.57 Mbit/s

Hosted by Interserver, inc (Secaucus, NJ) [5567.91 km]:
Testing download speed........................................
Download: 65.59 Mbit/s
Testing upload speed..................................................
Upload: 30.29 Mbit/s

Hosted by Vodafone DE (Frankfurt) [638.17 km]:
Testing download speed........................................
Download: 104.35 Mbit/s
Testing upload speed..................................................
Upload: 37.03 Mbit/s

Hosted by Vodafone NL (Utrecht) [366.45 km]:
Testing download speed........................................
Download: 38.35 Mbit/s
Testing upload speed..................................................
Upload: 31.97 Mbit/s

Ping to Google was super quick (likely because of the London Data Exchange (LDeX) network) and they have IPv6!

root@ldex:~/benchmark# ping -c 5 google.com
PING google.com (173.194.34.137) 56(84) bytes of data.
64 bytes from lhr14s21-in-f9.1e100.net (173.194.34.137): icmp_req=1 ttl=60 time=0.320 ms
64 bytes from lhr14s21-in-f9.1e100.net (173.194.34.137): icmp_req=2 ttl=60 time=0.399 ms
64 bytes from lhr14s21-in-f9.1e100.net (173.194.34.137): icmp_req=3 ttl=60 time=0.457 ms
64 bytes from lhr14s21-in-f9.1e100.net (173.194.34.137): icmp_req=4 ttl=60 time=0.497 ms
64 bytes from lhr14s21-in-f9.1e100.net (173.194.34.137): icmp_req=5 ttl=60 time=0.423 ms

--- google.com ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4004ms
rtt min/avg/max/mdev = 0.320/0.419/0.497/0.060 ms

root@ldex:~/benchmark# ping6 -c 5 google.com
PING google.com(lhr14s21-in-x05.1e100.net) 56 data bytes
64 bytes from lhr14s21-in-x05.1e100.net: icmp_seq=1 ttl=60 time=4.71 ms
64 bytes from lhr14s21-in-x05.1e100.net: icmp_seq=2 ttl=60 time=0.393 ms
64 bytes from lhr14s21-in-x05.1e100.net: icmp_seq=3 ttl=60 time=0.411 ms
64 bytes from lhr14s21-in-x05.1e100.net: icmp_seq=4 ttl=60 time=0.396 ms
64 bytes from lhr14s21-in-x05.1e100.net: icmp_seq=5 ttl=60 time=0.421 ms

--- google.com ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4005ms
rtt min/avg/max/mdev = 0.393/1.267/4.715/1.724 ms

Benchmark Tests

Disk IO is not the fastest but is very consistent. A friend on another node gets 160MB/s. 90MB/s is more than enough though for what most people use their vps for.

root@ldex:~# dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync; rm test
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 11.9295 s, 90.0 MB/s

ioping is consistent but is quite slow.

root@ldex:~/ioping# ./ioping -c 5 /root/
4096 bytes from /root/ (ext3 /dev/mapper/ldex-root): request=1 time=3.8 ms
4096 bytes from /root/ (ext3 /dev/mapper/ldex-root): request=2 time=3.8 ms
4096 bytes from /root/ (ext3 /dev/mapper/ldex-root): request=3 time=3.8 ms
4096 bytes from /root/ (ext3 /dev/mapper/ldex-root): request=4 time=3.8 ms
4096 bytes from /root/ (ext3 /dev/mapper/ldex-root): request=5 time=3.9 ms

--- /root/ (ext3 /dev/mapper/ldex-root) ioping statistics ---
5 requests completed in 4019.8 ms, 261 iops, 1.0 mb/s
min/avg/max/mdev = 3.8/3.8/3.9/0.0 ms

UnixBench results were great considering that this VPS only has one CPU core.

   #    #  #    #  #  #    #          #####   ######  #    #   ####   #    #
   #    #  ##   #  #   #  #           #    #  #       ##   #  #    #  #    #
   #    #  # #  #  #    ##            #####   #####   # #  #  #       ######
   #    #  #  # #  #    ##            #    #  #       #  # #  #       #    #
   #    #  #   ##  #   #  #           #    #  #       #   ##  #    #  #    #
    ####   #    #  #  #    #          #####   ######  #    #   ####   #    #

   Version 5.1.3                      Based on the Byte Magazine Unix Benchmark

   Multi-CPU version                  Version 5 revisions by Ian Smith,
                                      Sunnyvale, CA, USA
   January 13, 2011                   johantheghost at yahoo period com


1 x Dhrystone 2 using register variables  1 2 3 4 5 6 7 8 9 10

1 x Double-Precision Whetstone  1 2 3 4 5 6 7 8 9 10

1 x Execl Throughput  1 2 3

1 x File Copy 1024 bufsize 2000 maxblocks  1 2 3

1 x File Copy 256 bufsize 500 maxblocks  1 2 3

1 x File Copy 4096 bufsize 8000 maxblocks  1 2 3

1 x Pipe Throughput  1 2 3 4 5 6 7 8 9 10

1 x Pipe-based Context Switching  1 2 3 4 5 6 7 8 9 10

1 x Process Creation  1 2 3

1 x System Call Overhead  1 2 3 4 5 6 7 8 9 10

1 x Shell Scripts (1 concurrent)  1 2 3

1 x Shell Scripts (8 concurrent)  1 2 3

========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: ldex: GNU/Linux
   OS: GNU/Linux -- 2.6.32-5-686 -- #1 SMP Sun Sep 23 09:49:36 UTC 2012
   Machine: i686 (unknown)
   Language: en_US.utf8 (charmap="ANSI_X3.4-1968", collate="ANSI_X3.4-1968")
   CPU 0: QEMU Virtual CPU version (cpu64-rhel6) (4988.7 bogomips)
          x86-64, MMX, Physical Address Ext, SYSCALL/SYSRET
   21:46:08 up 26 min,  3 users,  load average: 0.26, 0.25, 0.10; runlevel 2

------------------------------------------------------------------------
Benchmark Run: Sat Feb 16 2013 21:46:08 - 22:14:08
1 CPU in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       16491551.9 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     2456.2 MWIPS (10.0 s, 7 samples)
Execl Throughput                               5646.9 lps   (29.5 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        677710.6 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          173440.1 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1934689.0 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1100679.6 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 309660.1 lps   (10.0 s, 7 samples)
Process Creation                              21220.6 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   7493.2 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    971.4 lpm   (60.0 s, 2 samples)
System Call Overhead                         943334.4 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   16491551.9   1413.2
Double-Precision Whetstone                       55.0       2456.2    446.6
Execl Throughput                                 43.0       5646.9   1313.2
File Copy 1024 bufsize 2000 maxblocks          3960.0     677710.6   1711.4
File Copy 256 bufsize 500 maxblocks            1655.0     173440.1   1048.0
File Copy 4096 bufsize 8000 maxblocks          5800.0    1934689.0   3335.7
Pipe Throughput                               12440.0    1100679.6    884.8
Pipe-based Context Switching                   4000.0     309660.1    774.2
Process Creation                                126.0      21220.6   1684.2
Shell Scripts (1 concurrent)                     42.4       7493.2   1767.3
Shell Scripts (8 concurrent)                      6.0        971.4   1619.1
System Call Overhead                          15000.0     943334.4    628.9
                                                                   ========
System Benchmarks Index Score                                        1214.4

You can view the Geekbench result online.

Conclusion:

CastleGem offer a stable and decent service. The LDEX network is impressive and their range of locations offers a lot of choice for their customers. Their staff have always been very friendly and helpful when we’ve had to communicate. Bernhard, the Managing Director, is very nice and seems to be focused on constantly improving CastleGem’s products and services. I love how you can customise the plan to your own suit when ordering. CastleGem are not primarily aimed at the LowEnd market but more the enterprise one. I believe this is reflected in the great service they provide and, unfortunately, the pricing. That said, their pricing is relatively low compared to what their competitors in Enterprise market charge and I believe it’s worth it.

Highly recommended.


Plans have recently been updated and now start from $12.50 [£8.00] /Quarter. Hopefully they’ll run another LowEnd offer sometime soon ;)

If you’re also a customer, please comment below! Thank you for reading, you can expect many more longterm reviews in the coming months.

(This review followed the same format as Max Schaefer’s earlier reviews.)

21 Comments

  1. Matthew Morgan:

    a vary good review.
    any info on what datacentre they use?

    February 17, 2013 @ 12:45 pm | Reply
  2. It is also worth pointing out that the plan in review is basically this one: https://worralorrasurfa.castlegem.co.uk/whmcs/cart.php?a=add&pid=100 at $17/Quarter ($5.60/month) (but with a free 256MB RAM upgrade). The above plan didn’t exist at the time I purchased this vps.

    February 17, 2013 @ 12:52 pm | Reply
  3. Pascal:

    Sounds good. Is this plan still somewhere available at $5.60/month?

    February 17, 2013 @ 2:02 pm | Reply
    • Pascal:

      (With 512MB RAM I mean)

      February 17, 2013 @ 2:11 pm | Reply
      • hello Pascal,

        as mentioned below, if you open a ticket (sales), we’ll be happy to sort something out for you if you refer to this thread :)

        thank you!
        oz/b.

        February 17, 2013 @ 4:17 pm | Reply
    • This was a customised deal I got when they had their old plans but it was provisioned with 512MB RAM instead of the 256MB I paid for. Basically this is a review of their 256Mb SATA mini with extra RAM.

      February 17, 2013 @ 2:53 pm | Reply
  4. Jack:

    Fantastic review! I will have to see what there network to Liverpool is like. I am with MiniVPS at the minute but always looking for another little vps.

    February 17, 2013 @ 3:07 pm | Reply
  5. Emil Vals:

    The support at Castlegem is also excellent.

    February 17, 2013 @ 3:16 pm | Reply
  6. Hello,

    first of all thank you for this review. I am very happy about the way it came out, highlighting both the pros and cons, giving us incentives to improve our service (even) further. For those of you who are interested in such plans, we are willing to make them available for the lowendbox/talk community if you open a sales ticket and refer to this review! :-)

    with gratitude, and thank you again, very much appreciate it!

    oz/bernard.

    February 17, 2013 @ 3:59 pm | Reply
  7. dano:

    Castlegem has been great to host with — I originally bought a 256 Xen VM last year sometime, and have since migrated over to the KVM offering. Support has been awesome, as they are friendly and fast, so I have zero complaints. The network is pretty good from most of EU and reaches India and Middle East pretty quickly also. The KVM VM I have now is snappy and is just as good or better than the Xen I had with them.

    Keep up the good work Castlegem.

    February 17, 2013 @ 5:05 pm | Reply
  8. Spirit:

    I am glad to see that you like them too Liam :) I haven’t tried CastleGem’s London location as I have more than a few VPSs around this area (country) already however VPS from Vienna (Austria – close to me) and Prague (Czech Republic – just slightly more away) works great for me.
    Another interesting location I see there is Ashburn, US – relatively rare US LEB location with great link between US and EU. If I would need another vps from .us for something more serious this would be my pick.

    February 17, 2013 @ 5:58 pm | Reply
    • thanks Spirit, as always, a pleasure! :)
      Ashburn was chosen for its routes to the EU and its location in the US, close to one of the major exchanges there. Has proven very reliable for people who need fast links to both NY and Chicago as well for example :)
      cheers –
      b.

      February 17, 2013 @ 6:40 pm | Reply
  9. May I know how the result below this line has been taken ??

    “For more accuracy, I ran a few more speedtests which were also good.”

    Below this line, some download and upload speed tests were done. May I know, how was it done ??

    February 17, 2013 @ 6:40 pm | Reply
  10. Liam is awesome!

    February 17, 2013 @ 11:58 pm | Reply
  11. awesome!

    February 18, 2013 @ 7:34 am | Reply
  12. m.sabouri:

    Is TUN/TAP enabled in their servers ?

    February 19, 2013 @ 4:32 am | 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. Some comments may be held temporarily for manual approval.
  • 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 *