All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
ClawCloud cheap 1vCore, 1 GB VPS and 2 vCore, 4 GB VDS benchmark & review
@CLAWCLOUD kindly provided access to the cheap 1 vCore, 1 GB mem. VPS as well as to the 2 vCores, 4 GB mem. VDS to me. Both are quite performant and I could complete > 50 runs within just 1.5 days and now tell you about what I experienced and saw.
Let me begin with one very positive and one negative news: the good one is that (I've been told) CC will support payment via PayPal very soon (early this year). Until then though payment is only possible via credit card or 2 chinese gateways (Ali-something), which might be quite unattractive for many western potential customers - but hey, PayPal will soon be available!
The bad news is that CC does not support ISOs but only a couple of linux distros (the usual ones) and IIRC neither Windows (Update/correction: they do offer Windows VDS but with a heavy "Windows tax"). Which obviously is a major disappointment for [some]BSD users as well as those who prefer some less common linux distro, meeh! And no, CC does not seem to have concrete plans, let alone a ETA to improve on that.
I had to put this front-up because it also explains why this benchmark was running on debian 12 rather than on FreeBSD, so take the disk performance numbers with a grain of salt (because linux is caching aggressively and making numbers look better).
Btw, another very, even extremely negative point that must be mentioned: I innocently opened a high prio ticket to ask them to please mount a FreeBSD ISO for me - I didn't get any reaction at all for over 24 hrs (at which point I stopped even looking for a response). That is absolutely not acceptable and makes CC look really bad, in particular as they are a new provider here.
Now to the benchmark, beginning with the small VPS, first as usual processor and memory.
Version 2.50a, (c) 2018+ jsg (->lowendtalk.com)
Machine: amd64, Arch.: x86_64, Model: Intel(R) Xeon(R) Platinum
[debian OS], version: Linux 6.1.0, Mem.: 923 MB
CPU - Cores: 1, Family/Model/Stepping: 6/85/4
Cache: 32K/32K L1d/L1i, 1024K L2, 33M L3
Std. Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
pse36 cflsh mmx fxsr sse sse2 ss htt sse3 pclmulqdq ssse3 fma cx16
pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline aes xsave osxsave
avx f16c rdrnd hypervisor
Ext. Flags: fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid pat pse36 rdseed
adx smap clflushopt clwb sha syscall nx pdpe1gb rdtscp lm lahf_lm
lzcnt
AES? Yes
Nested Virt.? No
HW RNG? Yes
ProcMem SC [MB/s]: avg 204.1 - min 84.6 (41.4 %), max 340.3 (166.7 %)
ProcMem MA [MB/s]: avg 307.7 - min 268.9 (87.4 %), max 332.9 (108.2 %)
ProcMem MB [MB/s]: avg 308.4 - min 270.4 (87.7 %), max 333.1 (108.0 %)
ProcMem AES [MB/s]: avg 912.3 - min 771.6 (84.6 %), max 1052.6 (115.4 %)
ProcMem RSA [kp/s]: avg 94.9 - min 84.6 (89.1 %), max 106.7 (112.4 %)
I like what I see there and the platinum scalable (Xeon), which largely boils down to "many cores", shows that one can get Epyc like performance, incl. with decent crypto hardware support. Spread is within normal limits; they obviously put plenty VPS on such an expensive processor but they don't overdo it.
On to the disk
--- Disk 4 KB - Buffered ---
Write seq. [MB/s]: avg 6.21 - min 4.52 (72.8%), max 7.43 (119.6%)
Write rnd. [MB/s]: avg 9.44 - min 6.92 (73.3%), max 12.07 (127.8%)
Read seq. [MB/s]: avg 606.51 - min 334.32 (55.1%), max 820.79 (135.3%)
Read rnd. [MB/s]: avg 1061.68 - min 151.96 (14.3%), max 1673.73 (157.6%)
--- Disk 4 KB - Sync/Direct ---
Write seq. [MB/s]: avg 5.81 - min 4.27 (73.5%), max 7.00 (120.4%)
Write rnd. [MB/s]: avg 8.04 - min 5.58 (69.4%), max 9.66 (120.1%)
Read seq. [MB/s]: avg 613.05 - min 343.90 (56.1%), max 829.31 (135.3%)
Read rnd. [MB/s]: avg 1091.07 - min 159.85 (14.7%), max 1784.07 (163.5%)
--- Disk 64 KB - Buffered ---
Write seq. [MB/s]: avg 64.86 - min 48.97 (75.5%), max 75.78 (116.8%)
Write rnd. [MB/s]: avg 92.54 - min 61.40 (66.3%), max 111.16 (120.1%)
Read seq. [MB/s]: avg 761.65 - min 575.04 (75.5%), max 851.56 (111.8%)
Read rnd. [MB/s]: avg 1490.51 - min 1058.59 (71.0%), max 1718.26 (115.3%)
--- Disk 64 KB - Sync/Direct ---
Write seq. [MB/s]: avg 62.72 - min 46.95 (74.9%), max 72.30 (115.3%)
Write rnd. [MB/s]: avg 80.43 - min 59.45 (73.9%), max 94.30 (117.2%)
Read seq. [MB/s]: avg 764.24 - min 586.77 (76.8%), max 873.51 (114.3%)
Read rnd. [MB/s]: avg 1499.74 - min 1066.48 (71.1%), max 1776.46 (118.5%)
--- Disk 1 MB - Buffered ---
Write seq. [MB/s]: avg 278.30 - min 223.86 (80.4%), max 310.31 (111.5%)
Write rnd. [MB/s]: avg 333.17 - min 269.89 (81.0%), max 362.23 (108.7%)
Read seq. [MB/s]: avg 304.48 - min 287.76 (94.5%), max 336.16 (110.4%)
Read rnd. [MB/s]: avg 1106.42 - min 999.52 (90.3%), max 1181.59 (106.8%)
--- Disk 1 MB - Sync/Direct ---
Write seq. [MB/s]: avg 277.63 - min 229.85 (82.8%), max 311.80 (112.3%)
Write rnd. [MB/s]: avg 296.95 - min 245.72 (82.7%), max 335.06 (112.8%)
Read seq. [MB/s]: avg 311.59 - min 304.56 (97.7%), max 339.50 (109.0%)
Read rnd. [MB/s]: avg 1104.92 - min 980.64 (88.8%), max 1190.90 (107.8%)
--- Disk IOps (Sync/Direct) ---
Write seq. [MB/s]: avg 16.06 - min 12.81 (79.8%), max 18.20 (113.3%)
IOps : avg 4111.61 - min 3278.10 (79.7%), max 4659.33 (113.3%)
Hmm, looks like SAS12 SSDs. The performance is not bad at all, but neither is it NVMe-like. 16 MB/s in 4k/4t and a bit over 4k IOps is nothing to write home about but neither is it badly slow.
My take is that for a cheap VPS that performance is really OK, certainly not a speed daemon to put a DB on but certainly good enough for most jobs LET users buy a cheap VPS for.
Finally the network/connectivity
--- Europe ---
NO OSL mirror.neuf.no [F: 0]
DL [Mb/s]: avg 228.9 - min 190.4 (83.2%), max 244.8 (106.9%)
Ping [ms]: avg 25.5 - min 25.4 (99.5%), max 26.4 (103.4%)
Web ping [ms]: avg 28.0 - min 26.3 (93.8%), max 54.6 (194.7%)
UK LON lon.speedtest.clouvider.net [F: 0]
DL [Mb/s]: avg 236.4 - min 213.9 (90.5%), max 249.8 (105.7%)
Ping [ms]: avg 17.1 - min 10.6 (62.0%), max 17.4 (101.8%)
Web ping [ms]: avg 17.7 - min 17.1 (96.8%), max 22.3 (126.3%)
NL AMS mirror.nl.leaseweb.net [F: 0]
DL [Mb/s]: avg 204.8 - min 42.0 (20.5%), max 260.3 (127.1%)
Ping [ms]: avg 8.5 - min 8.4 (99.2%), max 8.6 (101.6%)
Web ping [ms]: avg 16.9 - min 8.4 (49.6%), max 144.2 (851.4%)
DE FRA fra.lg.core-backbone.com [F: 0]
DL [Mb/s]: avg 254.5 - min 252.9 (99.4%), max 258.0 (101.4%)
Ping [ms]: avg 2.5 - min 2.3 (90.6%), max 2.7 (106.3%)
Web ping [ms]: avg 2.8 - min 2.4 (84.4%), max 3.4 (119.5%)
FR mirrors.ircam.fr [F: 0]
DL [Mb/s]: avg 227.4 - min 196.7 (86.5%), max 246.1 (108.3%)
Ping [ms]: avg 16.4 - min 15.5 (94.3%), max 19.7 (119.9%)
Web ping [ms]: avg 29.0 - min 16.4 (56.6%), max 76.2 (263.1%)
IT ROM debian.mirror.garr.it [F: 0]
DL [Mb/s]: avg 219.8 - min 211.4 (96.2%), max 240.3 (109.3%)
Ping [ms]: avg 25.2 - min 22.7 (90.0%), max 28.4 (112.6%)
Web ping [ms]: avg 52.0 - min 23.0 (44.3%), max 160.3 (308.4%)
RO BUC mirror.efect.ro [F: 0]
DL [Mb/s]: avg 234.6 - min 219.1 (93.4%), max 243.1 (103.6%)
Ping [ms]: avg 25.3 - min 25.1 (99.3%), max 25.6 (101.3%)
Web ping [ms]: avg 28.6 - min 25.3 (88.4%), max 49.2 (171.8%)
GR UNK ftp.otenet.gr [F: 4]
DL [Mb/s]: avg 199.3 - min 0.0 (0.0%), max 235.4 (118.1%)
Ping [ms]: avg 37.1 - min 0.0 (0.0%), max 41.3 (111.3%)
Web ping [ms]: avg 67.0 - min 0.0 (0.0%), max 810.9 (1210.1%)
--- Asia/Oceania ---
TR IST mirror.timtal.com.tr [F: 0]
DL [Mb/s]: avg 216.2 - min 205.9 (95.2%), max 235.3 (108.9%)
Ping [ms]: avg 38.1 - min 37.8 (99.2%), max 38.4 (100.8%)
Web ping [ms]: avg 40.7 - min 37.9 (93.1%), max 58.6 (143.9%)
RU MOS speedtest.hostkey.ru [F: 0]
DL [Mb/s]: avg 200.4 - min 176.5 (88.1%), max 230.3 (114.9%)
Ping [ms]: avg 36.1 - min 23.5 (65.1%), max 39.9 (110.6%)
Web ping [ms]: avg 48.9 - min 36.3 (74.2%), max 59.6 (121.8%)
IR ISF repo.iut.ac.ir [F: 0]
DL [Mb/s]: avg 158.1 - min 65.1 (41.2%), max 212.9 (134.7%)
Ping [ms]: avg 95.7 - min 0.0 (0.0%), max 131.7 (137.7%)
Web ping [ms]: avg 155.7 - min 0.0 (0.0%), max 403.9 (259.4%)
RU SIB mirror.truenetwork.ru [F: 0]
DL [Mb/s]: avg 168.6 - min 41.9 (24.9%), max 218.0 (129.3%)
Ping [ms]: avg 84.0 - min 55.6 (66.2%), max 88.7 (105.7%)
Web ping [ms]: avg 201.3 - min 84.7 (42.1%), max 617.8 (306.9%)
IN MU mirrors.piconets.webwerks.in [F: 0]
DL [Mb/s]: avg 157.6 - min 148.7 (94.4%), max 166.9 (105.9%)
Ping [ms]: avg 122.1 - min 121.6 (99.6%), max 122.4 (100.3%)
Web ping [ms]: avg 122.6 - min 121.6 (99.2%), max 127.4 (103.9%)
SG SGP mirror.sg.gs [F: 0]
DL [Mb/s]: avg 110.9 - min 80.5 (72.5%), max 125.1 (112.8%)
Ping [ms]: avg 154.9 - min 101.7 (65.7%), max 156.4 (101.0%)
Web ping [ms]: avg 170.0 - min 157.1 (92.4%), max 348.9 (205.2%)
CN HK mirrors.xtom.hk [F: 0]
DL [Mb/s]: avg 90.7 - min 46.9 (51.7%), max 106.3 (117.1%)
Ping [ms]: avg 227.2 - min 188.1 (82.8%), max 432.7 (190.4%)
Web ping [ms]: avg 249.1 - min 188.1 (75.5%), max 432.7 (173.7%)
JP TOK speedtest.tokyo2.linode.com [F: 1]
DL [Mb/s]: avg 70.6 - min 0.0 (0.0%), max 77.6 (109.9%)
Ping [ms]: avg 229.3 - min 0.0 (0.0%), max 246.0 (107.3%)
Web ping [ms]: avg 238.7 - min 0.0 (0.0%), max 261.5 (109.5%)
AU MEL ftp.au.freebsd.org [F: 0]
DL [Mb/s]: avg 72.2 - min 71.4 (98.9%), max 72.6 (100.6%)
Ping [ms]: avg 272.4 - min 269.9 (99.1%), max 277.3 (101.8%)
Web ping [ms]: avg 277.1 - min 271.8 (98.1%), max 333.6 (120.4%)
--- America (east coast to west coast)---
US MAS ftp5.freebsd.org [F: 0]
DL [Mb/s]: avg 92.9 - min 35.0 (37.7%), max 135.5 (145.8%)
Ping [ms]: avg 81.0 - min 80.7 (99.6%), max 84.7 (104.5%)
Web ping [ms]: avg 86.6 - min 80.8 (93.3%), max 102.0 (117.7%)
US WDC mirrors.vcea.wsu.edu [F: 0]
DL [Mb/s]: avg 118.0 - min 112.9 (95.6%), max 121.3 (102.8%)
Ping [ms]: avg 154.8 - min 154.7 (99.9%), max 155.7 (100.6%)
Web ping [ms]: avg 159.6 - min 157.9 (98.9%), max 163.7 (102.6%)
US DAL mirror.dal.nexril.net [F: 0]
DL [Mb/s]: avg 158.2 - min 154.2 (97.5%), max 163.4 (103.2%)
Ping [ms]: avg 119.7 - min 119.4 (99.7%), max 120.3 (100.5%)
Web ping [ms]: avg 120.7 - min 119.5 (99.0%), max 136.8 (113.3%)
US LAX la.speedtest.clouvider.net [F: 0]
DL [Mb/s]: avg 128.4 - min 123.2 (95.9%), max 137.6 (107.2%)
Ping [ms]: avg 138.9 - min 49.3 (35.5%), max 142.2 (102.4%)
Web ping [ms]: avg 146.1 - min 141.8 (97.1%), max 153.8 (105.3%)
US SFR mirrors.ocf.berkeley.edu [F: 0]
DL [Mb/s]: avg 125.0 - min 119.4 (95.5%), max 129.8 (103.9%)
Ping [ms]: avg 150.8 - min 121.4 (80.5%), max 152.3 (101.0%)
Web ping [ms]: avg 154.4 - min 151.6 (98.2%), max 174.9 (113.3%)
BR SAO ftp3.br.freebsd.org [F: 0]
DL [Mb/s]: avg 86.5 - min 83.7 (96.8%), max 88.1 (101.8%)
Ping [ms]: avg 227.5 - min 224.8 (98.8%), max 237.8 (104.5%)
Web ping [ms]: avg 249.6 - min 224.9 (90.1%), max 287.0 (115.0%)
--- Africa ---
KE UNK mirror.liquidtelecom.com [F: 0]
DL [Mb/s]: avg 155.2 - min 145.0 (93.4%), max 156.2 (100.7%)
Ping [ms]: avg 121.2 - min 0.0 (0.0%), max 125.6 (103.6%)
Web ping [ms]: avg 127.3 - min 0.0 (0.0%), max 179.5 (141.1%)
Yay, what a nice surprise! Those cheap VPS come (iirc) with a 200 Mb/s limit and they actually meet that spec and in fact even surpass it quite a few times. Nice! Even Istanbul (TR) and Moscow (RU) achieve about 200 Mb/s. Even Isfahan (IR), a known to be difficult to reach target achieves a very respectable 150+Mb/s and so does Mumbai (IN). Singapore a bit over and Hongkong (ĆN) a bit under 100 Mb/s also is uncommonly good. And Tokyo (JP) and Ozzyland (Melbourne) achieving about 70 Mb/s is very respectable as well. And even a target in Africa with a bit over 150 Mb/s!
I'm really impressed by those connectivity results!
TL;DR The disk is OK for a cheap VPS, the processor is almost too good for such a cheap VPS, and the connectivity borders miracle-land.
Strongly recommended on two conditions:
- (1) you get it at promo (super low) price
- (2) 1 Vcore, 1 GB mem. and a definitely not NVMe (but decent) SSD is good enough for you
For the normal (not promo) price I'd still recommend it if condition 2 is met -and- you absolutely need very good connectivity, especially to Asia.
Comments
OK, now the 2 vCores, 4 GB memory VDS.
First processor & memory
ProcMem RSA [kp/s]: avg 85.5 - min 79.9 (93.5 %), max 90.5 (105.9 %)
Pardon me, single core performance lower than the cheapie VPS and multi-core far below expectations and merely 50 or 60 MB/s more than single core?
On a VDS?
I'll generously assume that CC just has some early config hiccup, because if that really is what you get as a 2 vCore VDS then thanks no!
Now on to the disk
IOps : avg 4383.75 - min 3343.20 (76.3%), max 4737.00 (108.1%)
Meh, again hardly better than the cheap VPS. Nuff said.
On to the network
Yay, now we're talking! All european targets above 400 Mb/s - including even Istanbul (TR) and Moscow (RU) - most in fact even about 500 Mb/s and a few even above 1 Gb/s, DE, FRA even achieving almost 7 Gb/s, but thats within DE, FRA.
Iran and India above 150 Mb/s but strangely far east Asia (SGP and east of it) a bit worse than with the cheap VPS. Me surprised because the benchmakr runs happened at the same time (in parallel). Weird. But hey, still decent results.
The Americas otoh performed similarly or even somewhat better.
Summary
Unless you absolutely need very good connectivity within the VDS's region, not recommended, because the disk performance is roughly equal to that of the cheap VPS, the processor & mem. perform only slightly better. As VDS tend to be considerably more expensive than VPS you'll have way better bang per buck buying a VPS with more vCores and/or memory.
CloudClaw clearly should work both on their VDS and their product/price ranking.
TL;DR
The cheap VPS - which is really nice - with a 60% promo rebate costs about $17 per year, the VDS which, besides better regional connectivity, is only slightly more powerful however with the same 60% promo rebate costs about $67/yr - and you do not really get much more for the $50(!) difference and highly likely would be better off buying 3 additional cheapies.
Also note the rather low traffic volume! In Asia that may make sense but in Europe and NA it just makes the products look less attractive, sorry.
Worth buying
Yes, the response speed of support tickets is crucial! It's been three days since my last communication with ClowCloud regarding my high - priority support ticket, and I still haven't received a reply.
Please expect two weeks or longer. I don't know why, but they are just that slow.
Thank you so much @jsg for your objective review.
We’ll work on adding the custom ISO capability .
As we start to reach out to international users, your feedback on our product is extremely important. It guides us in setting the right technical priorities for the future.
ps.
The cheap VPS will be available for just $7/yr ,from January 9, 00:00 to January 23 EST (UTC-5)
1 vCPU (Xeon Platinum)
1 GB RAM
20 GB SSD
500 GB Bandwidth @ 200Mbps
1 IPv4/1 IPv6
$7/year
Are all regions participating in this event?
lg plz
link plz
Which areas are included?
for only one year or more than
hkg
japan
singpore
Thanks
8.209.70.255 https://lg.eu-central-1.claw.cloud/ - Frankfurt, Germany
47.253.157.212 https://lg.us-east-1.claw.cloud/ - N.Virginia, USA
47.251.75.195 https://lg.us-west-1.claw.cloud/ - N.California, USA
47.79.18.247 | 240b:4001:278:8402::ffff | https://lg.cn-hongkong-intl.claw.cloud/ |Hong Kong, China
47.79.32.210 | 240b:4009:25a:1801::ffff | https://lg.ap-northeast-1-intl.claw.cloud/ | Tokyo, Japan
47.79.122.185 | 240b:4000:93:de01::ffff | https://lg.ap-southeast-1-intl.claw.cloud/ | Singapore
hongkong ipv6 240b:4001:278:8402::ffff is cannot linked
You're welcome and I'm pleased to learn that my work was/is helpful
A very good question!
So, @CLAWCLOUD is that price recurring for those of us who get one within the promo period you described?
That's only for the first year.
Thanks, but I'd prefer to get an official answer from @CLAWCLOUD.
This special introductory price is available for the first year only. Afterward, renewals will be at the standard rate. Details of future renewal offers for existing customers are still to be determined.
that's a very comprehensive chart~
Will current customers be able to get one? And is it a limited quantity or flash deal?
Also: what's the difference between the new Cloud-VDS and China Optimized ones with your current offering of Dedicated VPS?
So how do we get this do we send a ticket or what
Try this: 240b:4001:278:8402::ffff
1.Yes, each user can get one, regardless of whether they are already a CLAWCLOUD customer.
2.Both series deliver outstanding performance, with the main difference being in their network routes.
The China-optimized series is specially tailored for accessing China, offering faster connection speeds and reliable performance with no packet loss, ensuring a smooth experience even during peak times.
hi @CLAWCLOUD can you confirm partnership/relationship with alibaba?
are you using their compute products or only network transit?
You opened a high priority ticket for that? I would have ignored it as well
Well, being able to install an OS IS high priority. And I meanwhile just so happen to know that their ignorance had an entirely different reason, not in any way linked specifically to this ticket.
it's 09.01.2025.
Forget performance. I want to try it. Where is the link?