Digital Ocean vs Linode

Null

Customer
I've just read @Matt's review of DigitalOcean and noticed a few things that are getting to me.

I've noticed something about the comparison chart that DO offer.

The 2GB ($20) DO 'Droplet' has a 40GB SSD and 3,000GB bandwidth.
The 2GB ($40) Linode has a 96GB HDD and 4,000GB bandwidth.

If you use 4,000GB of bandwidth with DO it'll cost you an additional $20 per month, which puts the DO droplet on par with the Linode box (in terms of pricing). When Linode move over to SSD's in the next month or so, you'll get an additional 46GB of data storage with Linode for the same price point.

On the point of SSD's; DO use pure SSD's and Linode will be using an SSD Hybrid solution. Which is better? In my opinion, Linode's SSD Hybrid solution. Why? Well, Linode are going to use the SSD's as a kind of cache, then your data will eventually hit the traditional HDD drives. If you run a forum, your most active files will be the ones served by the webserver, why put the other 95% of the systems files on a super fast SSD? They'll be touched every now and again, sure.
 

Matt

Owner
Another thing, the CPU. You get 2x2ghz CPU cores with DigitalOcean, but access to 8 CPU cores with Linode (2x priority), with the ability to burst to the full 8 cores.

Linode Unix Benchmark score
Code:
System Benchmarks Index Score                                        1333.1
DigitalOcean Unix Benchmark score
Code:
System Benchmarks Index Score                                         749.8
That's almost twice as fast for Linode!
 

Null

Customer
Another thing, the CPU. You get 2x2ghz CPU cores with DigitalOcean, but access to 8 CPU cores with Linode (2x priority), with the ability to burst to the full 8 cores.

Linode Unix Benchmark score
Code:
System Benchmarks Index Score                                        1333.1
DigitalOcean Unix Benchmark score
Code:
System Benchmarks Index Score                                         749.8
That's almost twice as fast for Linode!
Just imagine how much Linode's going to change in the next few weeks! DO has surely kicked Linode's arse into gear.
 

Matt

Owner
Yep, I'm really looking forward to their SSD offering, and will be spinning up a Linode as soon as I can to try it out.

It's good to see some healthy competition in the "Cloud" hosting game.
 

Matt

Owner
I've just finished comparing two $80 VPS setups from both Linode and DigitalOcean

DigitalOcean
Single CPU
System Benchmarks Index Score = 1029.7
Multi-Thread CPU
System Benchmarks Index Score = 1810.4

Linode
Single CPU
System Benchmarks Index Score = 373.6
Multi-Thread CPU
System Benchmarks Index Score 1365.2

That's quite a difference if you ask me!

$80 8GB DigitalOcean Droplet
Code:
[root@test ~]# free -m
             total       used       free     shared    buffers     cached
Mem:          7999        131       7867          0          4         28
-/+ buffers/cache:         98       7901
Swap:            0          0          0

Code:
[root@test ~]# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 2
model name      : QEMU Virtual CPU version 1.4.0
stepping        : 3
cpu MHz         : 1999.999
cache size      : 4096 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 4
wp              : yes
flags           : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm rep_good unfair_spinlock pni vmx cx16 popcnt hypervisor lahf_lm
bogomips        : 3999.99
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 6
model           : 2
model name      : QEMU Virtual CPU version 1.4.0
stepping        : 3
cpu MHz         : 1999.999
cache size      : 4096 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 4
wp              : yes
flags           : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm rep_good unfair_spinlock pni vmx cx16 popcnt hypervisor lahf_lm
bogomips        : 3999.99
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 2
vendor_id       : GenuineIntel
cpu family      : 6
model           : 2
model name      : QEMU Virtual CPU version 1.4.0
stepping        : 3
cpu MHz         : 1999.999
cache size      : 4096 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 4
wp              : yes
flags           : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm rep_good unfair_spinlock pni vmx cx16 popcnt hypervisor lahf_lm
bogomips        : 3999.99
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

processor       : 3
vendor_id       : GenuineIntel
cpu family      : 6
model           : 2
model name      : QEMU Virtual CPU version 1.4.0
stepping        : 3
cpu MHz         : 1999.999
cache size      : 4096 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 4
wp              : yes
flags           : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 clflush mmx fxsr sse sse2 syscall nx lm rep_good unfair_spinlock pni vmx cx16 popcnt hypervisor lahf_lm
bogomips        : 3999.99
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management:

[root@test ~]#

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

   System: test.mattwservices.co.uk: GNU/Linux
   OS: GNU/Linux -- 2.6.32-431.1.2.0.1.el6.x86_64 -- #1 SMP Fri Dec 13 13:06:13 UTC 2013
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: QEMU Virtual CPU version 1.4.0 (4000.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   CPU 1: QEMU Virtual CPU version 1.4.0 (4000.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   CPU 2: QEMU Virtual CPU version 1.4.0 (4000.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   CPU 3: QEMU Virtual CPU version 1.4.0 (4000.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   15:08:21 up 3 min,  2 users,  load average: 0.40, 0.19, 0.07; runlevel 3

------------------------------------------------------------------------
Benchmark Run: Mon Feb 17 2014 15:08:21 - 15:36:25
4 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       23953729.7 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     2664.5 MWIPS (9.6 s, 7 samples)
Execl Throughput                               2254.1 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        715144.5 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          198821.7 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1524713.6 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1301791.0 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 232133.5 lps   (10.0 s, 7 samples)
Process Creation                               5840.6 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   3132.8 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1186.8 lpm   (60.0 s, 2 samples)
System Call Overhead                        1743180.0 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   23953729.7   2052.6
Double-Precision Whetstone                       55.0       2664.5    484.5
Execl Throughput                                 43.0       2254.1    524.2
File Copy 1024 bufsize 2000 maxblocks          3960.0     715144.5   1805.9
File Copy 256 bufsize 500 maxblocks            1655.0     198821.7   1201.3
File Copy 4096 bufsize 8000 maxblocks          5800.0    1524713.6   2628.8
Pipe Throughput                               12440.0    1301791.0   1046.5
Pipe-based Context Switching                   4000.0     232133.5    580.3
Process Creation                                126.0       5840.6    463.5
Shell Scripts (1 concurrent)                     42.4       3132.8    738.9
Shell Scripts (8 concurrent)                      6.0       1186.8   1978.0
System Call Overhead                          15000.0    1743180.0   1162.1
                                                                   ========
System Benchmarks Index Score                                        1029.7

------------------------------------------------------------------------
Benchmark Run: Mon Feb 17 2014 15:36:25 - 16:04:25
4 CPUs in system; running 4 parallel copies of tests

Dhrystone 2 using register variables       52635115.3 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     8785.5 MWIPS (9.9 s, 7 samples)
Execl Throughput                               8593.8 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        374098.0 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          100665.8 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1114584.8 KBps  (30.0 s, 2 samples)
Pipe Throughput                             3079182.0 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 616848.1 lps   (10.0 s, 7 samples)
Process Creation                              27375.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   9868.3 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1359.9 lpm   (60.0 s, 2 samples)
System Call Overhead                        2672682.6 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   52635115.3   4510.3
Double-Precision Whetstone                       55.0       8785.5   1597.4
Execl Throughput                                 43.0       8593.8   1998.6
File Copy 1024 bufsize 2000 maxblocks          3960.0     374098.0    944.7
File Copy 256 bufsize 500 maxblocks            1655.0     100665.8    608.3
File Copy 4096 bufsize 8000 maxblocks          5800.0    1114584.8   1921.7
Pipe Throughput                               12440.0    3079182.0   2475.2
Pipe-based Context Switching                   4000.0     616848.1   1542.1
Process Creation                                126.0      27375.5   2172.7
Shell Scripts (1 concurrent)                     42.4       9868.3   2327.4
Shell Scripts (8 concurrent)                      6.0       1359.9   2266.5
System Call Overhead                          15000.0    2672682.6   1781.8
                                                                   ========
System Benchmarks Index Score                                        1810.4


4GB $80 Linode
Code:
[root@li689-151 ~]# free -m
             total       used       free     shared    buffers     cached
Mem:          3936        111       3825          0          2         20
-/+ buffers/cache:         88       3847
Swap:          511          0        511

Code:
[root@li689-151 ~]# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

processor       : 1
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

processor       : 2
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

processor       : 3
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

processor       : 4
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

processor       : 5
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

processor       : 6
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

processor       : 7
vendor_id       : GenuineIntel
cpu family      : 6
model           : 45
model name      : Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
stepping        : 7
microcode       : 0x70d
cpu MHz         : 2600.084
cache size      : 20480 KB
physical id     : 1
siblings        : 8
core id         : 7
cpu cores       : 1
apicid          : 47
initial apicid  : 47
fpu             : yes
fpu_exception   : yes
cpuid level     : 13
wp              : yes
flags           : fpu de tsc msr pae cx8 apic sep cmov pat clflush mmx fxsr sse sse2 ss ht syscall nx lm constant_tsc rep_good nopl nonstop_tsc pni pclmulqdq ssse3 cx16 sse4_1 sse4_2 popcnt tsc_deadline_timer aes hypervisor lahf_lm ida arat epb pln pts dtherm
bogomips        : 5202.84
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:

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

   System: li689-151: GNU/Linux
   OS: GNU/Linux -- 3.12.9-x86_64-linode37 -- #1 SMP Mon Feb 3 10:01:02 EST 2014
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 1: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 2: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 3: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 4: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 5: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 6: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 7: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (5202.8 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   20:31:11 up 5 min,  2 users,  load average: 0.86, 0.61, 0.27; runlevel 3

------------------------------------------------------------------------
Benchmark Run: Mon Feb 17 2014 20:31:11 - 20:59:52
8 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       18506496.5 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     2708.0 MWIPS (10.0 s, 7 samples)
Execl Throughput                               1039.3 lps   (29.8 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        180088.0 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           50073.4 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        605081.5 KBps  (30.0 s, 2 samples)
Pipe Throughput                              309583.9 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  21424.7 lps   (10.0 s, 7 samples)
Process Creation                               1930.2 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   2136.9 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    814.0 lpm   (60.0 s, 2 samples)
System Call Overhead                         293561.2 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   18506496.5   1585.8
Double-Precision Whetstone                       55.0       2708.0    492.4
Execl Throughput                                 43.0       1039.3    241.7
File Copy 1024 bufsize 2000 maxblocks          3960.0     180088.0    454.8
File Copy 256 bufsize 500 maxblocks            1655.0      50073.4    302.6
File Copy 4096 bufsize 8000 maxblocks          5800.0     605081.5   1043.2
Pipe Throughput                               12440.0     309583.9    248.9
Pipe-based Context Switching                   4000.0      21424.7     53.6
Process Creation                                126.0       1930.2    153.2
Shell Scripts (1 concurrent)                     42.4       2136.9    504.0
Shell Scripts (8 concurrent)                      6.0        814.0   1356.6
System Call Overhead                          15000.0     293561.2    195.7
                                                                   ========
System Benchmarks Index Score                                         373.6

------------------------------------------------------------------------
Benchmark Run: Mon Feb 17 2014 20:59:52 - 21:27:57
8 CPUs in system; running 8 parallel copies of tests

Dhrystone 2 using register variables      108831166.3 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                    17674.9 MWIPS (10.0 s, 7 samples)
Execl Throughput                               3803.1 lps   (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        258715.8 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           69262.6 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        796336.5 KBps  (30.0 s, 2 samples)
Pipe Throughput                             2121034.2 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 334803.7 lps   (10.0 s, 7 samples)
Process Creation                               9010.8 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   8175.4 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1085.7 lpm   (60.3 s, 2 samples)
System Call Overhead                        1774435.2 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0  108831166.3   9325.7
Double-Precision Whetstone                       55.0      17674.9   3213.6
Execl Throughput                                 43.0       3803.1    884.4
File Copy 1024 bufsize 2000 maxblocks          3960.0     258715.8    653.3
File Copy 256 bufsize 500 maxblocks            1655.0      69262.6    418.5
File Copy 4096 bufsize 8000 maxblocks          5800.0     796336.5   1373.0
Pipe Throughput                               12440.0    2121034.2   1705.0
Pipe-based Context Switching                   4000.0     334803.7    837.0
Process Creation                                126.0       9010.8    715.1
Shell Scripts (1 concurrent)                     42.4       8175.4   1928.2
Shell Scripts (8 concurrent)                      6.0       1085.7   1809.6
System Call Overhead                          15000.0    1774435.2   1183.0
                                                                   ========
System Benchmarks Index Score                                        1365.2
 

Matt

Owner
No.

The above test is a little unfair in that it's based upon pricing, not system resources.

If Matt attempts the same with an 8GB Linode, I'm sure Linode will come out on top.

@Matt, care to prove me wrong?
I'll do an 8GB linode tonight.
 

Matt

Owner
Spinning it up now. One thing to remember though, while comparing price points, it's also easy to forgot the same amount of RAM from Linode is now twice as expensive as DO, so depending on the results, most people won't want to pay 2x the amount per month for the same amount of RAM.
 

Matt

Owner
Just finished. Also span up a 4GB DigitalOcean droplet to cover the basic range off.

upload_2014-2-18_20-29-22.png

The 8GB ($160 a month!) Linode was a worse performer than the 4GB one, and performed worse overall than the $40 4GB DigitalOcean droplet. Past their 4GB mark, DigitalOcean wins hands down, especially when you consider it's also 50% cheaper than the same 8GB RAM allocation that Linode are offering.
 

Matt

Owner
Linode 8GB
Code:
========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: li417-57: GNU/Linux
   OS: GNU/Linux -- 3.12.9-x86_64-linode37 -- #1 SMP Mon Feb 3 10:01:02 EST 2014
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 1: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 2: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 3: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 4: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 5: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 6: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 7: Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (4535.5 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   19:10:41 up 5 min,  2 users,  load average: 0.26, 0.13, 0.05; runlevel 3

------------------------------------------------------------------------
Benchmark Run: Tue Feb 18 2014 19:10:41 - 19:38:42
8 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       18001286.1 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     2467.4 MWIPS (9.8 s, 7 samples)
Execl Throughput                               1107.7 lps   (29.8 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        220194.5 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           57711.6 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        651215.0 KBps  (30.0 s, 2 samples)
Pipe Throughput                              338706.3 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  33177.8 lps   (10.0 s, 7 samples)
Process Creation                               2285.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   2462.2 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    784.3 lpm   (60.1 s, 2 samples)
System Call Overhead                         315044.3 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   18001286.1   1542.5
Double-Precision Whetstone                       55.0       2467.4    448.6
Execl Throughput                                 43.0       1107.7    257.6
File Copy 1024 bufsize 2000 maxblocks          3960.0     220194.5    556.0
File Copy 256 bufsize 500 maxblocks            1655.0      57711.6    348.7
File Copy 4096 bufsize 8000 maxblocks          5800.0     651215.0   1122.8
Pipe Throughput                               12440.0     338706.3    272.3
Pipe-based Context Switching                   4000.0      33177.8     82.9
Process Creation                                126.0       2285.5    181.4
Shell Scripts (1 concurrent)                     42.4       2462.2    580.7
Shell Scripts (8 concurrent)                      6.0        784.3   1307.2
System Call Overhead                          15000.0     315044.3    210.0
                                                                   ========
System Benchmarks Index Score                                         414.0

------------------------------------------------------------------------
Benchmark Run: Tue Feb 18 2014 19:38:42 - 20:07:09
8 CPUs in system; running 8 parallel copies of tests

Dhrystone 2 using register variables       72079464.8 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                    13911.9 MWIPS (9.4 s, 7 samples)
Execl Throughput                               3741.2 lps   (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        191898.0 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           56543.2 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        550113.2 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1455891.4 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 228377.2 lps   (10.0 s, 7 samples)
Process Creation                               5643.9 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   6376.2 lpm   (60.1 s, 2 samples)
Shell Scripts (8 concurrent)                    847.4 lpm   (60.2 s, 2 samples)
System Call Overhead                        1584524.1 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   72079464.8   6176.5
Double-Precision Whetstone                       55.0      13911.9   2529.4
Execl Throughput                                 43.0       3741.2    870.0
File Copy 1024 bufsize 2000 maxblocks          3960.0     191898.0    484.6
File Copy 256 bufsize 500 maxblocks            1655.0      56543.2    341.7
File Copy 4096 bufsize 8000 maxblocks          5800.0     550113.2    948.5
Pipe Throughput                               12440.0    1455891.4   1170.3
Pipe-based Context Switching                   4000.0     228377.2    570.9
Process Creation                                126.0       5643.9    447.9
Shell Scripts (1 concurrent)                     42.4       6376.2   1503.8
Shell Scripts (8 concurrent)                      6.0        847.4   1412.4
System Call Overhead                          15000.0    1584524.1   1056.3
                                                                   ========
System Benchmarks Index Score                                        1030.5

DigitalOcean 4GB
Code:
========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: test.mattwservices.co.uk: GNU/Linux
   OS: GNU/Linux -- 2.6.32-431.1.2.0.1.el6.x86_64 -- #1 SMP Fri Dec 13 13:06:13 UTC 2013
   Machine: x86_64 (x86_64)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: QEMU Virtual CPU version 1.4.0 (4000.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   CPU 1: QEMU Virtual CPU version 1.4.0 (4000.0 bogomips)
          x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET, Intel virtualization
   14:31:58 up 3 min,  2 users,  load average: 0.12, 0.06, 0.01; runlevel 3

------------------------------------------------------------------------
Benchmark Run: Tue Feb 18 2014 14:31:58 - 14:59:59
2 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       21673002.1 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     2319.2 MWIPS (9.8 s, 7 samples)
Execl Throughput                               1862.2 lps   (29.8 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        456419.5 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          105461.5 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1052796.3 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1027970.7 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 141544.4 lps   (10.0 s, 7 samples)
Process Creation                               4160.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   2447.7 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    516.2 lpm   (60.1 s, 2 samples)
System Call Overhead                        1435172.6 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   21673002.1   1857.2
Double-Precision Whetstone                       55.0       2319.2    421.7
Execl Throughput                                 43.0       1862.2    433.1
File Copy 1024 bufsize 2000 maxblocks          3960.0     456419.5   1152.6
File Copy 256 bufsize 500 maxblocks            1655.0     105461.5    637.2
File Copy 4096 bufsize 8000 maxblocks          5800.0    1052796.3   1815.2
Pipe Throughput                               12440.0    1027970.7    826.3
Pipe-based Context Switching                   4000.0     141544.4    353.9
Process Creation                                126.0       4160.5    330.2
Shell Scripts (1 concurrent)                     42.4       2447.7    577.3
Shell Scripts (8 concurrent)                      6.0        516.2    860.4
System Call Overhead                          15000.0    1435172.6    956.8
                                                                   ========
System Benchmarks Index Score                                         724.0

------------------------------------------------------------------------
Benchmark Run: Tue Feb 18 2014 14:59:59 - 15:28:00
2 CPUs in system; running 2 parallel copies of tests

Dhrystone 2 using register variables       27539236.6 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     4126.5 MWIPS (9.9 s, 7 samples)
Execl Throughput                               3548.9 lps   (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        278684.9 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           91670.4 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        876578.8 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1600037.1 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 223537.3 lps   (10.0 s, 7 samples)
Process Creation                               9919.0 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   4528.3 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    639.1 lpm   (60.1 s, 2 samples)
System Call Overhead                        1955085.4 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   27539236.6   2359.8
Double-Precision Whetstone                       55.0       4126.5    750.3
Execl Throughput                                 43.0       3548.9    825.3
File Copy 1024 bufsize 2000 maxblocks          3960.0     278684.9    703.7
File Copy 256 bufsize 500 maxblocks            1655.0      91670.4    553.9
File Copy 4096 bufsize 8000 maxblocks          5800.0     876578.8   1511.3
Pipe Throughput                               12440.0    1600037.1   1286.2
Pipe-based Context Switching                   4000.0     223537.3    558.8
Process Creation                                126.0       9919.0    787.2
Shell Scripts (1 concurrent)                     42.4       4528.3   1068.0
Shell Scripts (8 concurrent)                      6.0        639.1   1065.1
System Call Overhead                          15000.0    1955085.4   1303.4
                                                                   ========
System Benchmarks Index Score                                         973.2
 

mikeg

Member
Great info! Thank you for making this available as I am currently evaluating both.
But aren't the results somewhat random since one person's account may be assigned to a lightly load server and someone else with same plan ends up on a very busy server?

I keep reading about the "noisy neighbor" effect and that neither company is very specific about the exact resources your receive:
https://forum.linode.com/viewtopic.php?t=10861&p=62689
"As hybinet brought up, the two servers may have different model CPUs. You can check that yourself with `cat /proc/cpuinfo`.
To see if noisy neighbors are limiting your CPU resources, run something CPU-intensive, open `top` or `htop`, and see if the "st" (steal) percentage is high."
https://www.digitalocean.com/community/questions/server-specs-1-core-equivalent-to
"Is there still no minimum guaranteed level of CPU performance? I could get all of a Xeon core or be totally starved?
For instance, I would like to know that my virtual CPUs will get at least x time slices on the CPU in every second, a definition of how small that slice was, and the specification of the lowest-end physical CPU in your cluster."

For my case, DigitalOcean's SSDs seem to outperform Linode, but I worry that DO is getting oversubscribed (or will be in the near future).

Thanks,

- mike
 

Null

Customer
Linode's hybrid SSD solution should be announced very soon.

From what I've been reading the past few hours, DO need to do some work on their drives, a lot of people are reporting drops in performance (across the net).
 

mikeg

Member
So darn, it would seem that the results are somewhat random -- I guess there is truth in YMMV with cloud/VPS hosting.

Next question: Are you using this UnixBench test: https://code.google.com/p/byte-unixbench/ ?
Of the test specific tests listed on that page, I haven't figured out which ones (if any) or most applicable to web hosting (maybe Process Creation and SystemCallOverhead? )
 

Matt

Owner
So darn, it would seem that the results are somewhat random -- I guess there is truth in YMMV with cloud/VPS hosting.

Next question: Are you using this UnixBench test: https://code.google.com/p/byte-unixbench/ ?
Of the test specific tests listed on that page, I haven't figured out which ones (if any) or most applicable to web hosting (maybe Process Creation and SystemCallOverhead? )
Yes, that is the test I was running.

The issues I've found specifically with Linode is the CPU throttling, and getting high levels is %st (steal time). I configured a 1GB VPS for @Null on here, and it took almost an hour to build and compile PHP! We re-provisioned the VPS on a 2GB plan to move onto a different node, and the issue wasn't there.

I had a lot of issues with %st when I was hosting my site on there, using a 2GB VPS as well. I had a smaller 1GB instance that I was using for Memcached and Elasticsearch, and that was actually faster than the 2GB in hosting terms.
 

mikeg

Member
Hi Matt, I guess your results pretty much confirm the "noisy neighbor" effect -- thank you for the data points.
In doing more research (wondering if I should just buy my own servers and colo again), I never realized that minor differences in hardware could also impact server performance to a significant degree. I would imagine the quote below also readily applies to VPS/Cloud servers:

http://stackoverflow.com/questions/...s-fast-on-a-2-53ghz-than-on-a-2-2ghz-core-2-d
In out CI environment we have a lot of "pizza box" computers that are supposed to be identical. They have the same hardware, were installed at the same time, and should be generally identical. They're even placed in "thermally equivalent" locations. They're not identical, and the variation can be quite stunning.

The only conclusion I have come up with is different binnings of CPU will have different thresholds for thermal stepping; some of the "best" chips hold up better. I also suspect other "minor" hardware faults/variations to be playing a role here. Maybe the slow boxes have slightly different components that play less well together ?
There are tools out there that will show you if your CPU is throttling for thermal reason
s.​

Interesting stuff that Vendor sales material never admits or addresses...
 

Matt

Owner
It certainly does make for an interesting read.

We've actually had an issue with disk speed on one of the linux boxes we have in the data centre at work, where the vibrations from the fans in the server below were causing issues read/write speeds.

Also, if airflow isn't exactly the same throughout the rack, that can also have an adverse effect. Someone stuck a router in a customers rack that we manage, and didn't rack it up properly, and it was getting heat soak one of the corners, which happened to be where the CPU was.....and it kept crapping out. Got an engineer to properly rack it up, and it's been running faultless for a few months now.

I've now moved my 2 busier sites and this site to the OVH server, and couldn't be happier at the minute. I'm actually paying less per month now on the dedicated box than I was on the last 3 VPS's that I had set up.

Same stack built with Centminmod:

upload_2014-3-8_15-11-55.png
 

Null

Customer
@mikeg, what size VPS are you looking for and what storage/transfer?

I'm working on a new project (with Matt) to offer VPS', I'd be more than happy to further discuss your needs.
 

mikeg

Member
We've actually had an issue with disk speed on one of the linux boxes we have in the data centre at work, where the vibrations from the fans in the server below were causing issues read/write speeds.
Yeah, I remember those days. At my last startup ('95 -2007) we bought & owned all the servers and built our own data room, so rack density wasn't really an issue and the boxes always had lots of room breath (but maintaing A/C equipment was another story).

In my research on OVH, I found some not so nice comments about their DDoS approach, but people always complain about stuff, so not sure what to think:
http://www.webhostingtalk.com/showthread.php?t=1269181
 
Top