Switching it up on y’all this month. By request by one of my dear readers, I’ve
included Vultr’s new “high frequency” server in this month’s comparison.
Traditionally these posts have been all the same price point to keep things
apples to apples, but for a buck more per month, it’s a pretty trivial price
difference, plus I was curious to see how well it performed!
Does this open the door for reviewing lower price points like the $3.50
instance from Lightsail? Maybe so.
That said, I’m actually working towards doing more expansive reviews that cover
not only more providers (feel like a new one pops up every day) but also a wide
range of plans and pricing.
Since last month, Linode dropped new GPU instances that are $1000 per month and
I can’t wait to give those a run too.
Vultr also has a new interface that was quite snappy, especially compared to
Linode and UpCloud which still don’t let you create multiple instances in one
shot.
As always, I spun up 3 server instances with each provider and averaged the
results. Each was running Ubuntu 18.04 LTS and I tried to pick data centers
closest to the New York area since that’s the Vultr data center that supported
the new high frequency instances.
Overview
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Location | New York 1 | Virginia, Zone A | Newark, NJ | Chicago 1 | New York (NJ) |
RAM | 1 GB | 1 GB | 1 GB | 1 GB | 1 GB |
CPU | 1 Core | 1 Core | 1 Core | 1 Core | 1 Core |
Storage | 25 GB SSD | 40 GB SSD | 25 GB SSD | 25 GB SSD | 32 GB NVMe |
Transfer | 1 TB | 2 TB | 1 TB | 1 TB | 1 TB |
Base Price | $5/month | $5/month | $5/month | $5/month | $6/month |
Backups | $1/month | N/A | $2/month | $0.06/GB | $1.20/month |
Transfer Overage | $0.01/GB | $0.09/GB | $0.02/GB | $0.056/GB | $0.01/GB |
Load Balancer | $10/month | $18/month | $10/month | N/A | N/A |
CPU Info
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
CPU MHz | 2294.61 | 2398.21 | 2500.00 | 2999.99 | 3792.00 |
Cache Size (KB) | 25344.00 | 30720.00 | 16384.00 | 16384.00 | 16384.00 |
BogoMips | 3059.67 | 3200.33 | 3333.00 | 3999.67 | 5056.33 |
CPU
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Events per Second | 962.57 | 862.13 | 671.88 | 1056.61 | 1326.61 |
Minimum (ms) | 0.92 | 1.05 | 1.24 | 0.91 | 0.74 |
Average (ms) | 1.05 | 1.16 | 1.55 | 0.95 | 0.75 |
Maximum (ms) | 4.76 | 1.80 | 20.54 | 3.04 | 2.29 |
Memory (Read)
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Ops per Second | 3826077.84 | 826451.83 | 2570694.61 | 4408247.54 | 5389794.26 |
Minimum (ms) | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
Average (ms) | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
Maximum (ms) | 1.83 | 1.90 | 19.81 | 2.93 | 1.55 |
Memory (Write)
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Ops per Second | 3846968.16 | 825357.64 | 2694374.78 | 4396213.81 | 5413351.61 |
Minimum (ms) | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
Average (ms) | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
Maximum (ms) | 2.99 | 1.08 | 16.31 | 2.42 | 1.44 |
File I/O
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Reads per Second | 1019.39 | 1301.29 | 1022.06 | 3689.44 | 4985.58 |
Writes per Second | 679.57 | 867.54 | 681.34 | 2459.60 | 3323.70 |
Fsyncs per Second | 2170.19 | 2770.91 | 2170.38 | 7861.80 | 10630.49 |
Minimum (ms) | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
Average (ms) | 0.29 | 0.20 | 0.62 | 0.07 | 0.05 |
Maximum (ms) | 31.89 | 11.60 | 53.30 | 7.46 | 3.26 |
MySQL
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Transactions per Second | 1711.33 | 2214.67 | 1679.67 | 4241.67 | 6371.33 |
Queries per Second | 34226.67 | 44293.33 | 33593.33 | 84833.33 | 127426.67 |
Minimum (ms) | 2.03 | 2.42 | 3.04 | 1.62 | 1.21 |
Average (ms) | 6.51 | 4.53 | 16.72 | 2.36 | 1.57 |
Maximum (ms) | 282.77 | 36.26 | 385.04 | 25.84 | 14.35 |
Redis
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
PING_INLINE | 55943.32 | 56402.87 | 32284.92 | 66879.25 | 63438.60 |
PING_BULK | 54206.55 | 55942.85 | 34044.87 | 66026.28 | 63079.04 |
SET | 54051.97 | 56378.98 | 30405.91 | 66882.39 | 63803.63 |
GET | 50844.48 | 56075.45 | 29490.34 | 66630.40 | 63859.51 |
INCR | 53050.26 | 56232.50 | 33640.56 | 66463.77 | 63763.35 |
LPUSH | 53264.27 | 54808.84 | 34887.06 | 67269.99 | 64427.61 |
RPUSH | 54746.99 | 55583.56 | 31433.06 | 67289.54 | 64735.00 |
LPOP | 53956.98 | 55645.72 | 31790.55 | 65941.98 | 64346.10 |
RPOP | 56712.86 | 56005.77 | 32003.94 | 64679.27 | 64593.60 |
SADD | 55955.49 | 56413.27 | 32421.09 | 64699.20 | 64805.88 |
HSET | 52579.00 | 54530.07 | 28868.29 | 66008.63 | 64951.40 |
SPOP | 52497.87 | 56204.79 | 33743.38 | 65132.67 | 64388.30 |
LRANGE_100 (first 100 elements) | 29121.21 | 30178.07 | 17174.31 | 33329.59 | 39304.99 |
LRANGE_300 (first 300 elements) | 12905.37 | 11734.52 | 7077.29 | 13209.16 | 17160.01 |
LRANGE_500 (first 500 elements) | 8559.31 | 8226.72 | 4722.25 | 9204.87 | 12477.33 |
LRANGE_600 (first 600 elements) | 6912.31 | 6556.58 | 3560.55 | 7189.39 | 9866.45 |
MSET (10 keys) | 48366.78 | 40839.24 | 30123.81 | 46870.44 | 56327.62 |
Speed Test
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Distance (km) | 2435.00 | 2090.30 | 2397.72 | 1575.67 | 2415.42 |
Latency (ms) | 45.650 | 80.049 | 49.842 | 41.387 | 48.373 |
Download (Mbit/s) | 1380.05 | 163.61 | 81.59 | 147.67 | 1428.46 |
Upload (Mbit/s) | 462.93 | 49.09 | 321.36 | 463.87 | 565.73 |
Apache Benchmark (against nginx
on the servers)
DigitalOcean | Lightsail | Linode | UpCloud | Vultr | |
---|---|---|---|---|---|
Requests per Second | 179.28 | 163.24 | 182.63 | 194.69 | 175.79 |
Time per Request (ms) (mean) | 2809.96 | 3082.49 | 2738.75 | 2568.92 | 2844.48 |
Transfer Rate (Kbyte/sec) | 150.39 | 136.93 | 153.20 | 163.32 | 147.46 |
Conclusion
With the exception of the Redis benchmarks and the Apache benchmarks (which are
always somewhat anecdotal), the new high frequency instances from Vultr
absolutely crushed it in every other category.
For only a buck more per month, Vultr’s new high frequency instances get you
more storage, CPU clock speed, raw performance as well as Vultr’s speedy network
(which seems to be back to peak performance).
As always, these benchmarks don’t necessarily simulate real world situations.
It’s always good to weigh out your own technical needs and pick a host that
matches that.
If you found this post helpful, it would be greatly appreciated if you used one
of my referral links below.