Wetin Be SSD VPS, and Wetin Flash Storage Do

Wetin be SSD VPS?

SSD VPS na virtual private server wey dem disk back am with flash memory instead of spinning hard disk. SSD mean solid state drive: storage wey NAND flash chips build, and e no get moving parts inside. VPS mean virtual private server: one isolated part of physical host machine, wey dey run im own operating system, and dem sell am to you like say na complete server. When you join the two together, the label dey promise one thing. When your server read data block, no mechanical part need move first.

Na the full definition be that. Everything wey follow dey explain wetin the label no talk, because the words "SSD hosting" for pricing page no tell you much about the storage array wey dey behind am.

If you still dey figure out how VPS different from ordinary virtual machine, the difference between VPS, VM and VPC na the better page to read first.

Hosts storage speed advertise dey instead of storage size

Plan page go list CPU cores, memory, disk size, and bandwidth, then e add one word about the disk wey no be size at all. Hosts dey do this because capacity stop to be the important storage number many years ago. The two numbers wey decide how server go feel na IOPS (input/output operations per second) and latency (how long one operation take before e return).

Na the mechanical difference cause am. Hard disk dey keep data for rotating platters and dey read am with head wey dey move for arm. To reach block for another place on platter, the arm must seek, then disk must turn until the correct sector reach under the head. For 7200 revolutions per minute, half turn dey take about 4 ms on average, and seek dey add several more. Flash no get arm or platter, so read na electrical lookup wey dey return within tens of microseconds.

[
  {
    "device": "7200 rpm hard disk",
    "random_read_iops": "125",
    "read_latency_ms": 8
  },
  {
    "device": "SATA SSD",
    "random_read_iops": "90,000",
    "read_latency_ms": 0.15
  },
  {
    "device": "NVMe SSD",
    "random_read_iops": "600,000",
    "read_latency_ms": 0.08
  }
]

7200 rpm hard disk get rating of roughly 125 random 4k reads per second, and e dey take about 8 ms for one read to return. NVMe drive get rating close to 600,000 of the same reads, at about 0.08 ms each. SATA SSD wey dey between dem get rating close to 90,000. Read these numbers as orders of magnitude, no be percentage improvement.

Two warnings dey about this data. These na published specifications for complete drives of each class, so dem be manufacturer sheet figures, no be measurements wey dem take for VPS. And you no go ever get complete drive: your volume na share of one device, or share of an array, wey dey beside other customers for the same hardware.

The two columns answer different questions too, so read dem together. Latency na how long you dey wait for one operation. IOPS na how many operations the drive fit process at the same time. Flash dey reach high IOPS figures through parallelism, because many flash chips dey answer many requests at once while deep queue dey keep dem busy. Single-threaded program wey issue one read, wait for am, then issue the next one no go ever see the highest figure for that chart. E go see the latency column instead.

SSD, NVMe, SATA and PCIe: four okwu for four different layers

Buyers dey mix these words up because each one dey describe different part of the system.

• SSD na the medium. E mean say data dey sit for NAND flash chips instead of magnetic platters.
• SATA na interface wey dem design during mechanical disk era. E reach maximum of 6 Gbit/s, wey be around 550 MB/s real throughput, and e get command queue wey fit hold 32 outstanding commands.
• NVMe (non-volatile memory express) na protocol wey dem write specially for flash. E support many queues wey each fit hold thousands of commands, so several CPU cores fit talk to the drive at the same time without sharing one narrow queue.
• PCIe (peripheral component interconnect express) na the bus wey NVMe dey run on, the same kind of lanes wey graphics card dey plug into.

So, SATA SSD and NVMe SSD both dey store data for flash. The difference na the interface wey each one dey use. SATA SSD still dey much faster than any hard disk, but that 32 command queue dey limit how much parallel work e fit handle, and parallel work na exactly wetin flash dey good at. Which one worth paying for depend on your workload, and the NVMe versus SATA SSD decision explain the matter properly.

Local flash or network attached storage?

Dem arrangements different well-well, but dem dey sell both under the same name.

Local storage mean say the flash drives dey inside the same physical host wey your VPS dey use. Request dey travel through PCIe inside one machine and come back directly. Na so latency fit remain for tens of microseconds.

Network attached storage mean say your virtual disk dey for separate storage cluster, often Ceph or SAN (storage area network), and every read and write dey cross network before e reach am. Providers usually call am "cloud block storage" or "elastic volumes". The flash na real one. But the network trip na real too, and dem add am to every operation. So latency fit reach high hundreds of microseconds or low milliseconds, instead of tens of microseconds.

Neither one wrong. Network storage fit survive if host die, because the data never dey for that host. Provider fit start your server for another hardware, and the disk go follow am. Local NVMe dey faster, but e dey tied to one physical machine. If hardware fail there, you go need restore from backup. Ask which one the plan dey use. Almost nobody dey ask.

How the difference dey feel for real server

Server work mostly na small random reads and writes, no be long sequential transfers. Na why the MB/s figure for marketing copy na the least useful number for the page.

• Database commits. Database wey promise durability calls fsync when transaction commits, then e wait for drive to confirm say the data really store. For hard disk, that wait dey take milliseconds, and e limit small database to low hundreds of commits per second. For flash, the same wait na fraction of one millisecond. Na here the gap widest, whether you dey run PostgreSQL, MySQL or SQLite as production database.
• Package installs. apt install dey unpack thousands of small files and sync dem to disk as e dey work. Almost none of this work na sequential, so IOPS na wetin dey limit am.
• Container image pulls. docker pull dey fetch compressed layers through network, then extract dem into thousands of small files. Network dey limit the download. Disk dey limit the extraction, and for slow volume, na extraction be the part wey you go sit down wait for.
• Boot and reboot. Startup dey read kernel and initramfs, then hundreds of small unit files and shared libraries wey scatter across the volume.

None of this na large sequential read. Volume wey dey stream 500 MB/s but dey serve only 3,000 IOPS go still feel slow during a docker compose pull, because dem dey count the wait per file, no be per megabyte.

Why “SSD cloud hosting” for pricing page no dey tell you almost anything

The word just describe the storage medium and stop there. E no tell you anything about the interface wey the drive dey use, and e no tell you whether the drive dey for the same machine as your server. E still no tell you the maximum level wey your plan fit reach.

That maximum level na the thing wey matter pass, but providers hardly advertise am. Providers dey limit IOPS and throughput for each volume, because one host dey serve many customers and one neighbour wey no get limit fit use all the resources and leave the rest with nothing. Limit of some thousand IOPS for hardware wey fit handle hundreds of thousands na normal and honest, but plan description no dey show am. Two plans fit both say “SSD”, while one na local NVMe with no per volume limit and the other na shared cluster volume wey dem limit to 3,000 IOPS.

Limits dey come in two forms. Sustained limit na fixed maximum wey no dey change. Burst limit give you small baseline plus credits wey allow you pass the limit for some time, and the credits dey refill when the volume idle. Burst limit fit look very good for five minute test, then e go drop to the baseline while database import or large restore dey happen. If provider quote large number, ask how long you fit maintain am.

How to check wetin your VPS really give you

From inside the guest, you fit only see wetin hypervisor tell you.

lsblk -d -o NAME,ROTA,MODEL,SIZE
cat /sys/block/vda/queue/rotational

Replace vda with the device name lsblk wey dem print for your disk. ROTA and the rotational file read 0 when kernel don receive information say the device no dey rotate, and 1 when dem tell am the opposite. Virtual disk dey set that flag based on wetin hypervisor advertise, so e describe the virtual device, no be the physical hardware underneath am. MODEL dey usually empty for virtio disk like vda, or e fit be generic string like QEMU HARDDISK for emulated SATA controller. Guests no dey meant to see the host array, and dem no dey see am.

So treat the flag as hint and measure the remaining things. Run random 4k test with fio against file for the volume, with the queue depth wey your application really dey use, and make e run long enough to use all burst credit. How to benchmark VPS properly get the fio commands and common mistakes wey fit produce numbers wey look better than reality.

Question three wey you suppose ask provider before you buy

1. Storage dey local to hypervisor, or na network-attached storage? The answer go set the minimum latency wey you fit expect, and e go determine wetin go happen to your data if physical host fail. Provider wey answer this one clearly don think about the matter.
2. Wetin be the IOPS limit for my volume? Ask for exact number. “Unlimited” and “enterprise grade” no be numbers. If no cap really dey, ask wetin go stop neighbour wey dey the same host from using the whole array during their backup window.
3. That limit dey sustained, or na burst limit? If na burst, ask for the baseline and how long the burst fit last. The baseline na the figure wey your nightly job go work with, so na around am you suppose plan.

Flash dey wear out, and na your problem?

Flash cells fit only take limited number of write cycles, so drives dey publish endurance rating for TBW (terabytes written) or DWPD (drive writes per day). The drive dey spread writes evenly across its cells; dem dey call this wear levelling. E also keep spare blocks in reserve to replace cells wey fail. For VPS, na provider problem be the wear: dem dey monitor SMART counters for their own drives and replace hardware before the rating finish. Your own problem older pass flash. Drive no be backup, and redundancy under your volume no be backup too, because e go copy deletion just as faithfully as e copy data.

The real trade-off wey you dey make na price per gigabyte. Flash cost more per gigabyte than spinning disks, so SSD plan usually get less capacity than hard disk plan for the same price. If you need plenty space for media or archives, keep the fast volume small and put the bulk for somewhere cheaper. This na the correct arrangement for keeping backups off the server. To see how storage fit inside the remaining bill, what a VPS actually costs dey break down the line items.

FAQ

SSD VPS na the same thing as NVMe VPS?

Every NVMe VPS na SSD VPS, because NVMe drives na flash. But SSD VPS no always be NVMe VPS. Plan wey dem advertise as "SSD" fit use SATA SSD. SATA SSD na flash wey dey behind interface wey dem design for mechanical disks, with 32 command queue and speed limit near 550 MB/s. Both dey much faster than hard disk. If the difference matter for your workload, ask the provider which one the plan dey use. No rely only on the plan name.

SSD VPS go make my website faster?

E go make disk work faster, and na only disk work. Page wey dey run several database queries for every request go faster, because the queries and commits na small random I/O. Page wey dey serve from memory or cache no touch disk as e dey go out, so e no go change much. Measure which part of the request slow before you pay for storage to fix am.

How I fit check whether my VPS really dey use SSD storage?

From inside the guest, you no fit verify the physical hardware. lsblk -d -o NAME,ROTA shows wetin the virtual device advertise, and hypervisor dey decide that value, so 0 there na hint, no be proof. The practical check na measurement: run fio with random 4k read workload for several minutes, then check the latency wey e report. Single digit millisecond random reads fit mean spinning disks or congested network volume. Tens of microseconds fit mean local flash.

Network attached SSD storage worse pass local NVMe?

E slower for each operation, and e dey fail in different way. Network round trip dey add to every read and write, so latency dey higher even though both na flash. In exchange, your data no dey for one physical host, so host failure no go carry the volume along, and snapshots plus live migration dey easier for provider. Choose local flash for database wey latency matter. Choose network storage when keeping the volume available matter pass microseconds.