joshtronic

Posted in VPS Showdown

VPS Showdown - October 2024 - DigitalOcean vs. Linode vs. Vultr

This post is old news – Get fresh benchmarks at VPS Showdown (daily updates!)

Time for the spooky season edition of the VPS Showdown. The last few months have been a lot of the same, so I decided to take another crack at trying to do an apples to apples comparison, focusing on price.

Even though DigitalOcean lacks a $5 plan, as the plans increase, they eventually offer a plan that's the same price as comparable offerings from Linode and Vultr. Turns out that Vultr actually has multiple plans that fit into the price point, so I had to make a judgement call on which one to pick.

Honestly starting to think that Vultr may deserve a dedicated post comparing the different plans that they have at the same price point.

So this month, I went with 4 GB instances, that are $24 per month. A bit pricier, but this tends to be the size that I pick for most things I have in production in a VPS. I usually will run things on a dedicated CPU when available, but for this month, it's all shared computing.

DigitalOcean and Linode are both on their standard shared offerings. For Vultr, I ended up going with the Premium AMD option. They also had Premium Intel and High Velocity (which had significantly more storage) plans. I could have went with a lesser option, but the price would have been less, and I didn't want that.

As per usual, I went ahead and spun up three instances with each provider, all running Ubuntu 24.04 LTS and living in the New York / Newark area. Where applicable, I've averaged the results.

Overview - DigitalOcean vs. Linode vs. Vultr

Version
Linux Distro Ubuntu 24.04 LTS
Kernel Version 6.8.0
MySQL Version 8.0.39
Redis Version 7.0.15

| | DigitalOcean | Linode | Vultr | | Location | New York 1 | Newark, NJ | New York (NJ) | | Monthly Price | $24/month | $24/month | $24/month | | Hourly Price | $0.036/hour | $0.036/hour | $0.033/hour | | RAM | 4 GB | 4 GB | 4 GB | | CPU | 2 Cores | 2 Cores | 2 Cores | | Storage | 80 GB | 80 GB | 100 GB | | Storage Type | SSD | SSD | NVMe | | Transfer | 4 TB | 4 TB | 5 TB | | Transfer Overage | $0.01/GB | $0.01/GB | $0.01/GB | | Backups | Starting at $4.80/month | $5/month | $4.80/month |

CPU Info

| | Model Name | | DigitalOcean - All Instances | DO-Regular | | Linode - Instance #1 | AMD EPYC 7601 32-Core Processor | | Linode - Instance #2 | AMD EPYC 7642 48-Core Processor | | Linode - Instance #3 | AMD EPYC 7542 32-Core Processor | | Vultr - All Instances | AMD EPYC-Rome Processor |

| | DigitalOcean | Linode | Vultr | | CPU MHz | 2,294.61 | 2,466.66 | 1,996.25 | | Cache Size (KB) | 4,096.00 | 512.00 | 512.00 | | BogoMips | NaN | 3,400.00 | 2,661.67 |

CPU

| | DigitalOcean | Linode | Vultr | | Events per Second | 800.29 | 1,472.68 | 3,328.02 | | Minimum (ms) | 0.98 | 0.64 | 0.28 | | Average (ms) | 1.25 | 0.69 | 0.30 | | Maximum (ms) | 7.70 | 4.41 | 8.71 |

Memory

Read

| | DigitalOcean | Linode | Vultr | | Ops per Second | 3,235,935.41 | 4,775,000.04 | 5,051,266.61 | | Minimum (ms) | 0.00 | 0.00 | 0.00 | | Average (ms) | 0.00 | 0.00 | 0.00 | | Maximum (ms) | 1.26 | 3.53 | 2.92 |

Write

| | DigitalOcean | Linode | Vultr | | Ops per Second | 3,019,862.82 | 4,695,861.71 | 5,004,443.75 | | Minimum (ms) | 0.00 | 0.00 | 0.00 | | Average (ms) | 0.00 | 0.00 | 0.00 | | Maximum (ms) | 1.57 | 4.67 | 2.37 |

File I/O

| | DigitalOcean | Linode | Vultr | | Reads per Second | 831.13 | 4,743.62 | 2,995.44 | | Writes per Second | 554.07 | 3,162.41 | 1,996.96 | | Fsyncs per Second | 1,777.02 | 10,127.38 | 6,396.65 | | Minimum (ms) | 0.00 | 0.00 | 0.00 | | Average (ms) | 0.32 | 0.06 | 0.09 | | Maximum (ms) | 62.26 | 21.19 | 34.44 |

MySQL

Read Only

| | DigitalOcean | Linode | Vultr | | Transactions per Second | 3,017.67 | 5,420.67 | 5,032.00 | | Queries per Second | 48,282.67 | 86,730.67 | 80,512.00 | | Minimum (ms) | 1.66 | 1.51 | 1.49 | | Average (ms) | 3.50 | 1.98 | 2.03 | | Maximum (ms) | 22.68 | 6.34 | 14.80 |

Write Only

| | DigitalOcean | Linode | Vultr | | Transactions per Second | 488.00 | 5,015.67 | 3,489.00 | | Queries per Second | 2,928.00 | 30,094.00 | 20,934.00 | | Minimum (ms) | 8.00 | 1.12 | 1.31 | | Average (ms) | 21.28 | 2.10 | 3.16 | | Maximum (ms) | 283.92 | 27.37 | 92.00 |

Read Write

| | DigitalOcean | Linode | Vultr | | Transactions per Second | 548.67 | 2,488.00 | 1,835.67 | | Queries per Second | 10,973.33 | 49,760.00 | 36,713.33 | | Minimum (ms) | 8.91 | 2.90 | 3.08 | | Average (ms) | 20.33 | 4.25 | 5.86 | | Maximum (ms) | 111.92 | 16.64 | 71.96 |

Redis

| | DigitalOcean | Linode | Vultr | | PING_INLINE | 42,969.42 | 55,466.76 | 50,586.94 | | PING_MBULK | 48,197.47 | 54,221.34 | 50,433.65 | | SET | 51,347.75 | 55,836.52 | 49,359.50 | | GET | 51,622.23 | 56,731.93 | 50,647.70 | | INCR | 52,018.66 | 54,653.47 | 50,124.57 | | LPUSH | 51,566.23 | 54,196.44 | 48,809.00 | | RPUSH | 51,804.00 | 57,714.46 | 49,045.15 | | LPOP | 52,300.89 | 57,295.52 | 51,371.91 | | RPOP | 54,025.06 | 56,388.18 | 50,525.94 | | SADD | 48,948.34 | 56,664.62 | 50,749.76 | | HSET | 46,012.93 | 56,203.15 | 50,779.56 | | SPOP | 51,673.22 | 57,871.55 | 52,257.89 | | LRANGE_100 (first 100 elements) | 26,198.40 | 42,881.93 | 40,662.65 | | LRANGE_300 (first 300 elements) | 12,235.53 | 22,875.56 | 23,733.62 | | LRANGE_500 (first 500 elements) | 7,115.77 | 16,397.17 | 17,183.65 | | LRANGE_600 (first 600 elements) | 6,917.01 | 14,705.01 | 15,285.05 | | MSET (10 keys) | 47,102.01 | 60,489.79 | 55,524.29 |

Conclusion

I was expecting that Vultr would clean up this month, with the advantage of a premium AMD processor. I kinda forgot that Linode offers AMD processors with all of their plans, so Vultr lost their edge a bit.

Overall Linode and Vultr's performance was quite comparable. DigitalOcean paled in comparison, but raw performance metrics don't tell the entire tale. Especially since DigitalOcean offers a ton more in terms of managed services, and their App Platform.

If you have found these posts informative and helpful in searching for a new hosting provider, please consider using one of the links below when signing up: