I think there are many students who don’t have money and want to use VPS as a hobby.
I was also looking for a free VPS to activate my license and to run my own WordPress domain.
In Japan, GCP’s GCE (Google Compute Engine) is famous, but there are other free VPSs as well.
In this article, I have researched the free VPS in the world and share them with you.
Google – GCE (Google Compute Engine)
The most famous free VPS is probably GCE.
The specifications are as follows.
E2-Micro
- CPU: 0.25 vCPU
- RAM: 1GB
- Storage: 30GB
Probably 30GB storage is enough for most users.
The vCPU stands for virtual CPU, 0.25vCPU means that 25% of 1vCPU can be used.
I benchmarked them with UnixBench and here are the results.
========================================================================
BYTE UNIX Benchmarks (Version 5.1.3)
System: instance-1: GNU/Linux
OS: GNU/Linux -- 3.10.0-1160.42.2.el7.x86_64 -- #1 SMP Tue Sep 7 14:49:57 UTC 2021
Machine: x86_64 (x86_64)
Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
CPU 0: Intel(R) Xeon(R) CPU @ 2.20GHz (4400.5 bogomips)
Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
CPU 1: Intel(R) Xeon(R) CPU @ 2.20GHz (4400.5 bogomips)
Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
17:42:47 up 5 min, 1 user, load average: 0.28, 0.18, 0.07; runlevel 2021-10-11
------------------------------------------------------------------------
Benchmark Run: Mon Oct 11 2021 17:42:47 - 17:49:44
2 CPUs in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 6815861.8 lps (10.0 s, 1 samples)
Double-Precision Whetstone 513.5 MWIPS (11.6 s, 1 samples)
Execl Throughput 300.4 lps (29.8 s, 1 samples)
File Copy 1024 bufsize 2000 maxblocks 56222.0 KBps (30.2 s, 1 samples)
File Copy 256 bufsize 500 maxblocks 14369.0 KBps (30.0 s, 1 samples)
File Copy 4096 bufsize 8000 maxblocks 162969.0 KBps (30.0 s, 1 samples)
Pipe Throughput 86890.4 lps (10.0 s, 1 samples)
Pipe-based Context Switching 6413.1 lps (10.0 s, 1 samples)
Process Creation 790.9 lps (30.2 s, 1 samples)
Shell Scripts (1 concurrent) 545.9 lpm (60.0 s, 1 samples)
Shell Scripts (8 concurrent) 94.2 lpm (60.5 s, 1 samples)
System Call Overhead 72114.7 lps (10.0 s, 1 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 6815861.8 584.0
Double-Precision Whetstone 55.0 513.5 93.4
Execl Throughput 43.0 300.4 69.9
File Copy 1024 bufsize 2000 maxblocks 3960.0 56222.0 142.0
File Copy 256 bufsize 500 maxblocks 1655.0 14369.0 86.8
File Copy 4096 bufsize 8000 maxblocks 5800.0 162969.0 281.0
Pipe Throughput 12440.0 86890.4 69.8
Pipe-based Context Switching 4000.0 6413.1 16.0
Process Creation 126.0 790.9 62.8
Shell Scripts (1 concurrent) 42.4 545.9 128.8
Shell Scripts (8 concurrent) 6.0 94.2 157.0
System Call Overhead 15000.0 72114.7 48.1
========
System Benchmarks Index Score 99.1
------------------------------------------------------------------------
Benchmark Run: Mon Oct 11 2021 17:49:44 - 17:57:04
2 CPUs in system; running 2 parallel copies of tests
Dhrystone 2 using register variables 5753240.4 lps (10.2 s, 1 samples)
Double-Precision Whetstone 882.1 MWIPS (44.3 s, 1 samples)
Execl Throughput 482.2 lps (30.1 s, 1 samples)
File Copy 1024 bufsize 2000 maxblocks 72794.0 KBps (30.0 s, 1 samples)
File Copy 256 bufsize 500 maxblocks 19576.0 KBps (30.0 s, 1 samples)
File Copy 4096 bufsize 8000 maxblocks 205287.0 KBps (30.0 s, 1 samples)
Pipe Throughput 127013.8 lps (10.0 s, 1 samples)
Pipe-based Context Switching 25768.3 lps (10.2 s, 1 samples)
Process Creation 1666.9 lps (30.0 s, 1 samples)
Shell Scripts (1 concurrent) 700.9 lpm (60.0 s, 1 samples)
Shell Scripts (8 concurrent) 94.2 lpm (60.5 s, 1 samples)
System Call Overhead 102356.7 lps (10.0 s, 1 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 5753240.4 493.0
Double-Precision Whetstone 55.0 882.1 160.4
Execl Throughput 43.0 482.2 112.1
File Copy 1024 bufsize 2000 maxblocks 3960.0 72794.0 183.8
File Copy 256 bufsize 500 maxblocks 1655.0 19576.0 118.3
File Copy 4096 bufsize 8000 maxblocks 5800.0 205287.0 353.9
Pipe Throughput 12440.0 127013.8 102.1
Pipe-based Context Switching 4000.0 25768.3 64.4
Process Creation 126.0 1666.9 132.3
Shell Scripts (1 concurrent) 42.4 700.9 165.3
Shell Scripts (8 concurrent) 6.0 94.2 156.9
System Call Overhead 15000.0 102356.7 68.2
========
System Benchmarks Index Score 147.4
The System Benchmarks Index Score is the overall score.
The CPUs are two Intel Xeon @ 2.20GHz. (As I explained earlier, two CPUs are not fully available).
Advantages
The network is 1Gbps, so it is relatively fast.
It is also the most famous VPS, so it is relatively easy to find out how to use it, which may be a good point for beginners.
Disadvantages
SMTP is not available because Port 25, 465 and 587 are not available.
Oracle – OCI (Oracle Cloud Infrastructure)
This is a VPS provided by Oracle.
The specification is as follows.
VM.Standard.E2.1.Micro
- CPU: AMD EPYC 7551 (2.0GHz-3.0GHz)、またはAMD EPYC 7742(2.25GHz-3.4GHz)、1/8 OCPU
- RAM: 1GB
- Storage: 50GB + 150GB (Block volume)
VM.Standard.A1.Flex
- CPU: Ampere Altra Q80-30 (-3.0GHz)、最大4OCPU
- RAM: Up to 24GB
Flex is clearly more powerful, but due to chronic resource shortage, I have not had a chance to use it yet.
The benchmark results for VM.Standard.E2.1.Micro are as follows
========================================================================
BYTE UNIX Benchmarks (Version 5.1.3)
System: mail-server: GNU/Linux
OS: GNU/Linux -- 5.4.17-2102.203.6.el7uek.x86_64 -- #2 SMP Wed Jul 21 17:51:54 PDT 2021
Machine: x86_64 (x86_64)
Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
CPU 0: AMD EPYC 7551 32-Core Processor (3992.5 bogomips)
Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
CPU 1: AMD EPYC 7551 32-Core Processor (3992.5 bogomips)
Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
13:08:58 up 20 days, 2:07, 1 user, load average: 0.27, 0.87, 0.69; runlevel 2021-09-22
------------------------------------------------------------------------
Benchmark Run: Tue Oct 12 2021 13:08:58 - 13:15:42
2 CPUs in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 13601372.2 lps (10.0 s, 1 samples)
Double-Precision Whetstone 1978.4 MWIPS (10.3 s, 1 samples)
Execl Throughput 1461.0 lps (29.6 s, 1 samples)
File Copy 1024 bufsize 2000 maxblocks 297889.0 KBps (30.0 s, 1 samples)
File Copy 256 bufsize 500 maxblocks 82898.0 KBps (30.0 s, 1 samples)
File Copy 4096 bufsize 8000 maxblocks 748380.0 KBps (30.0 s, 1 samples)
Pipe Throughput 457607.6 lps (10.0 s, 1 samples)
Pipe-based Context Switching 13813.2 lps (10.0 s, 1 samples)
Process Creation 1962.9 lps (30.0 s, 1 samples)
Shell Scripts (1 concurrent) 1719.9 lpm (60.1 s, 1 samples)
Shell Scripts (8 concurrent) 273.2 lpm (60.2 s, 1 samples)
System Call Overhead 700103.6 lps (10.0 s, 1 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 13601372.2 1165.5
Double-Precision Whetstone 55.0 1978.4 359.7
Execl Throughput 43.0 1461.0 339.8
File Copy 1024 bufsize 2000 maxblocks 3960.0 297889.0 752.2
File Copy 256 bufsize 500 maxblocks 1655.0 82898.0 500.9
File Copy 4096 bufsize 8000 maxblocks 5800.0 748380.0 1290.3
Pipe Throughput 12440.0 457607.6 367.9
Pipe-based Context Switching 4000.0 13813.2 34.5
Process Creation 126.0 1962.9 155.8
Shell Scripts (1 concurrent) 42.4 1719.9 405.6
Shell Scripts (8 concurrent) 6.0 273.2 455.3
System Call Overhead 15000.0 700103.6 466.7
========
System Benchmarks Index Score 388.5
------------------------------------------------------------------------
Benchmark Run: Tue Oct 12 2021 13:15:42 - 13:22:26
2 CPUs in system; running 2 parallel copies of tests
Dhrystone 2 using register variables 13925188.8 lps (10.1 s, 1 samples)
Double-Precision Whetstone 1962.2 MWIPS (10.1 s, 1 samples)
Execl Throughput 1360.2 lps (29.2 s, 1 samples)
File Copy 1024 bufsize 2000 maxblocks 264182.0 KBps (30.0 s, 1 samples)
File Copy 256 bufsize 500 maxblocks 70607.0 KBps (30.0 s, 1 samples)
File Copy 4096 bufsize 8000 maxblocks 630849.0 KBps (30.0 s, 1 samples)
Pipe Throughput 482708.5 lps (10.1 s, 1 samples)
Pipe-based Context Switching 72464.5 lps (10.0 s, 1 samples)
Process Creation 3020.8 lps (30.0 s, 1 samples)
Shell Scripts (1 concurrent) 1882.4 lpm (60.0 s, 1 samples)
Shell Scripts (8 concurrent) 258.9 lpm (60.3 s, 1 samples)
System Call Overhead 646748.6 lps (10.1 s, 1 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 13925188.8 1193.2
Double-Precision Whetstone 55.0 1962.2 356.8
Execl Throughput 43.0 1360.2 316.3
File Copy 1024 bufsize 2000 maxblocks 3960.0 264182.0 667.1
File Copy 256 bufsize 500 maxblocks 1655.0 70607.0 426.6
File Copy 4096 bufsize 8000 maxblocks 5800.0 630849.0 1087.7
Pipe Throughput 12440.0 482708.5 388.0
Pipe-based Context Switching 4000.0 72464.5 181.2
Process Creation 126.0 3020.8 239.7
Shell Scripts (1 concurrent) 42.4 1882.4 444.0
Shell Scripts (8 concurrent) 6.0 258.9 431.5
System Call Overhead 15000.0 646748.6 431.2
========
System Benchmarks Index Score 443.6
With an overall score of 443.6, it is three times faster than GCE.
It is not simply comparable, but as a machine specification OCI seems to be by far the better choice.
Advantages
It has a relatively high machine specification and a high limit on the number of machines.
E2.1.Micro is free for up to 2 machines and VM.Standard.A1.Flex is free for up to 4 machines with the minimum configuration (1OCPU, 6GB).
VM.Standard.A1.Flex is a configuration that would cost you more than 50 USD if you were to sign up for it normally.
This is more than most users can handle, and the fact that it’s free is a huge advantage.
Disadvantages
As with GCE, SMTP is not available for new users.
As for communication with the outside world, it seems to be limited to 50 Mbps, which I have to say is slow. (Although the limit may be different for VM.Standard.A1.Flex).
Flex is very difficult to secure. (especially in Tokyo region).
Other free VPS
There are other free VPSs available, but I would not recommend them in terms of reliability.
Conclusion
Basically, OCI is the way to go.
The specifications of VM.Standard.A1.Flex are more than enough for most users.
As for VM.Standard.E2.1.Micro, its machine specification surpasses that of the GCE.
As long as you are aware of the 1GB of memory, you will be able to use it in practical use.
Extra
Regarding OCI’s 50Mbps limit, it is possible to operate without being aware of the limit to some extent by using a CDN such as Cloudflare.
Also, Cloudflare hides your IP address and protects your website from various attacks such as DDoS.
Of course, it is necessary to take measures on the VPS side, but the barrier of introduction is low, so I recommend you to introduce Cloudflare.