All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Is this benchmark good?
I am having a server with
2x Intel Deca-Core E5-2670v2 - 128GB - 480GB SSD
I am not satisfied with the server as it running too slow, It took 90 min to install HestiaCP which normal get install in 3-5 in a low-end-vps. Have been complain to the dedi-provider that something is wrong with my server but they clam every is good.
`
Basic System Information:
Uptime : 0 days, 1 hours, 18 minutes
Processor : Intel(R) Xeon(R) CPU E5-2670 v2 @ 2.50GHz
CPU cores : 40 @ 1200.000 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 125.8 GiB
Swap : 4.0 GiB
Disk : 434.9 GiB
Distro : Ubuntu 22.04.2 LTS
Kernel : 5.15.0-69-generic
VM Type : NONE
Net Online : IPv4
IPv4 Network Information:
Location : North Kansas City, Missouri (MO)
Country : United States
fio Disk Speed Tests (Mixed R/W 50/50):
| Block Size | 4k (IOPS) | 64k (IOPS) |
|---|---|---|
| Read | 496.00 KB/s (124) | 6.93 MB/s (108) |
| Write | 523.00 KB/s (130) | 7.29 MB/s (113) |
| Total | 1.01 MB/s (254) | 14.23 MB/s (221) |
| Block Size | 512k (IOPS) | 1m (IOPS) |
| ------ | --- ---- | ---- ---- |
| Read | 31.20 MB/s (60) | 37.13 MB/s (36) |
| Write | 33.25 MB/s (64) | 40.21 MB/s (39) |
| Total | 64.45 MB/s (124) | 77.34 MB/s (75) |
iperf3 Network Speed Tests (IPv4):
| Provider | Location (Link) | Send Speed | Recv Speed | Ping |
|---|---|---|---|---|
| Clouvider | London, UK (10G) | 870 Mbits/sec | 866 Mbits/sec | 111 ms |
| Scaleway | Paris, FR (10G) | busy | 565 Mbits/sec | 105 ms |
| NovoServe | North Holland, NL (40G) | 875 Mbits/sec | 736 Mbits/sec | 104 ms |
| Uztelecom | Tashkent, UZ (10G) | busy | 610 Mbits/sec | 202 ms |
| Clouvider | NYC, NY, US (10G) | 914 Mbits/sec | 924 Mbits/sec | 41.9 ms |
| Clouvider | Dallas, TX, US (10G) | 938 Mbits/sec | 939 Mbits/sec | 10.6 ms |
| Clouvider | Los Angeles, CA, US (10G) | 920 Mbits/sec | 923 Mbits/sec | 40.7 ms |
Geekbench 6 Benchmark Test:
Test | Value
|
Single Core | 554
Multi Core | 5587
Full Test | https://browser.geekbench.com/v6/cpu/720924
`
Do benchmark looks good for the this type of serve?
Comments
likely a faulty or dogshit ssd
other than that, it looks fine
They claim that they changed it with brand new Kingston SSD.
The FIO is very low, especially for SSD, even HDD could be higher than that, and the GB6 Score also very low.
Smells like amateur Romania host. Run.
>
Yes, I am going to cancel server with them today and move to a better provider may be AMD Ryzen™ 9 5950X or equal.
SSD might have all it's pages in use etc, make sure discard is enabled AND do a fstrim -av to discard unused blocks.
I've seen that stuff happen so many times that typically i set aside 1-10% of an SSD in any server just so that it can internally do all that stuff easier and retain it's performance.
Once SSD gets full they are dogshit slow -- easily solved by discarding unused blocks / discarding whole device if you can / secure erase.
Also some SSDs are just garbage. especially old ones
Your metrics are lower than typical 7200rpm HDD, by a big margin.
I backup my data and reinstalled the whole OS again with latest OS Ubuntu 22.04 and did test again and that was the result, And when i complain the provide they said it performing good.
https://prnt.sc/lyw0UeZ1VGyo
Terrible read/write speeds. Are you sure it's an SSD?
You should publish the providers name...
Your network is good. But not with your storage.
Disk performance is bang on for HDD. Even dramless sata ssd (expected when they quote Kingston brand) should not be that bad
Your SSD performance is too bad. You need to ask them for replacement with nVME drives that will dramatically improve your app performance.
Just share the below performance with your provider that SSD IOPS is too bad compared to SSD IOPS below. It is from the 34Euro server.
Run smartctl and tell us what model those drives are
reinstall does not wipe the SSD completely, it has to be a DISCARD event of some sort. Writing over old data does not erase old data.
do a blkdiscard which wipes everything, and then issue reinstall if the host has system for remote reboots & reinstalls, otherwise ask help from the host.
Better yet, boot into rescue mode, and you can see swiftly the changes.
Doesn't matter, every provider has faulty hardware, or issues like this, only question is how often.
Publishing on forums could just be that they sold a lot with used SSDs without blkdiscard in between and default templates do not have discard enabled.
Hardware breaks and/or has issues no matter who owns it, not even Chuck Norris can intimidate the hardware not to break
Agrees but not accepting the issues is why i asked to post.
Its true hardware breaks sometimes, But even sending them proof that disk is so slow/broken, they are saying this is expected performance and that is disappointing.
Do the discard / fstrim like i told you instead of arguing against.
If it does not perform any better at that point, open ticket with your provider that you suspect the drive is broken // in a state of poor performance you are unable to fix, are they able to help? Remember to be friendly.
Who are "they"? Disclose bogus provider and provide redacted ticket chat screenshots where we could see both sides interacting.
Youngsters these days... Blablabla, without substantially backing their claims.
wow... 496.00 KB/s