All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
multi-location benchmark and review of the orangevps promo
While I'm personally settled wrt Asia and USA VPS, unless an extraordinarily tasty and performant (as in "connectivity") promo comes up, I still was triggered by the current @orangevps promo. IF those VPS are performant and offer a good bang per buck ratio, that is ...
So I asked them and they graciously provided me with access to all three VPS/locations and even for two weeks! THANK YOU, orangevps!
Before I present numbers I first want to mention something: their panel is one of the most comfortable ones I ever experienced. It looks a bit different (mostly a question of personal taste I guess) but it works flawlessly. In particular installing from ISO, which often is a problem or at least cumbersome, just works, period. Even better their VNC is really well set up. Unlike with most providers where I need to first find out which keyboard layout they happen to have configured (usually a US one) and then have to keep diverse key locations in mind and often struggle (e.g. imagine you're in vi, and are done and now need the bloody ':' key in order to save and quit ...), with the orangevps it almost feels surreal; you just type and it somehow (well, theoretically that information is available but most panels don't give a flying fuck) just works. Almost magical.
Kudos orangevps, you've done an exceptionally good job with your panel!
Being at kudos, their typical support reaction time seems to be about only three hours.
Now let's get at the beef, first the HongKong VPS. And btw. all three VPSs did at least 50 runs.
Version 2.5.0a, (c) 2018+ jsg (->lowendtalk.com)
Machine: amd64, Arch.: amd64, Model: Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz
OS, version: FreeBSD 14.2, Mem.: 3.989 GB
CPU - Cores: 2, Family/Model/Stepping: 6/85/7
Cache: 32K/32K L1d/L1i, 2M L2, 16M 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 sse3 pclmulqdq vmx ssse3 fma cx16
pdcm 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 mpx pat pse36
rdseed adx smap clflushopt clwb sha umip pku ospke syscall nx pdpe1gb
rdtscp lm lahf_lm lzcnt
AES? Yes
InNested Virt.? Yes
HW RNG? Yes
ProcMem SC [MB/s]: avg 228.3 - min 90.8 (39.8 %), max 371.8 (162.8 %)
ProcMem MA [MB/s]: avg 632.0 - min 522.0 (82.6 %), max 689.6 (109.1 %)
ProcMem MB [MB/s]: avg 651.6 - min 535.1 (82.1 %), max 713.0 (109.4 %)
ProcMem AES [MB/s]: avg 1101.5 - min 1091.7 (99.1 %), max 1106.2 (100.4 %)
ProcMem RSA [kp/s]: avg 95.4 - min 90.8 (95.2 %), max 100.4 (105.3 %)
Hmm, a Xeon Platinum, nice. And indeed decent performance and even decent crypto results!
So far I like what I see.
Is the disk a good performer as well? Let's see
--- Disk 4 KB - Buffered ---
Write seq. [MB/s]: avg 2.35 - min 2.19 (93.2%), max 2.66 (113.2%)
Write rnd. [MB/s]: avg 2.89 - min 2.72 (94.1%), max 3.18 (110.0%)
Read seq. [MB/s]: avg 14.42 - min 12.94 (89.7%), max 16.60 (115.1%)
Read rnd. [MB/s]: avg 13.67 - min 11.43 (83.6%), max 15.81 (115.7%)
--- Disk 4 KB - Sync/Direct ---
Write seq. [MB/s]: avg 2.33 - min 2.23 (95.7%), max 2.46 (105.5%)
Write rnd. [MB/s]: avg 2.85 - min 2.70 (94.6%), max 2.99 (104.7%)
Read seq. [MB/s]: avg 14.27 - min 11.87 (83.2%), max 15.29 (107.2%)
Read rnd. [MB/s]: avg 13.35 - min 8.06 (60.4%), max 16.03 (120.1%)
--- Disk 64 KB - Buffered ---
Write seq. [MB/s]: avg 99.71 - min 85.65 (85.9%), max 109.03 (109.3%)
Write rnd. [MB/s]: avg 68.59 - min 64.15 (93.5%), max 74.22 (108.2%)
Read seq. [MB/s]: avg 2731.83 - min 1825.28 (66.8%), max 3966.01 (145.2%)
Read rnd. [MB/s]: avg 153.23 - min 140.53 (91.7%), max 175.21 (114.3%)
--- Disk 64 KB - Sync/Direct ---
Write seq. [MB/s]: avg 11.03 - min 9.25 (83.8%), max 12.69 (115.0%)
Write rnd. [MB/s]: avg 5.54 - min 4.55 (82.1%), max 6.40 (115.5%)
Read seq. [MB/s]: avg 2504.54 - min 1791.07 (71.5%), max 4157.18 (166.0%)
Read rnd. [MB/s]: avg 152.80 - min 140.73 (92.1%), max 177.35 (116.1%)
--- Disk 1 MB - Buffered ---
Write seq. [MB/s]: avg 147.12 - min 123.22 (83.8%), max 161.98 (110.1%)
Write rnd. [MB/s]: avg 249.09 - min 219.45 (88.1%), max 273.63 (109.9%)
Read seq. [MB/s]: avg 3886.89 - min 2708.14 (69.7%), max 4465.29 (114.9%)
Read rnd. [MB/s]: avg 694.92 - min 630.57 (90.7%), max 770.61 (110.9%)
--- Disk 1 MB - Sync/Direct ---
Write seq. [MB/s]: avg 33.50 - min 27.12 (80.9%), max 36.71 (109.6%)
Write rnd. [MB/s]: avg 22.12 - min 17.76 (80.3%), max 24.24 (109.6%)
Read seq. [MB/s]: avg 3793.20 - min 2754.36 (72.6%), max 4446.40 (117.2%)
Read rnd. [MB/s]: avg 688.05 - min 630.30 (91.6%), max 739.69 (107.5%)
--- Disk IOps (Sync/Direct) ---
Write seq. [MB/s]: avg 9.38 - min 7.93 (84.6%), max 9.57 (102.0%)
IOps : avg 2400.72 - min 2031.36 (84.6%), max 2449.38 (102.0%)
Nuh, not really. It's OK I guess but not what one would like for a dynamic website and/or even medium DB work loads.
Maybe connectivity will save the day?
--- Europe ---
NO OSL mirror.terrahost.no [F: 1]
DL [Mb/s]: avg 53.6 - min 0.0 (0.0%), max 55.5 (103.6%)
Ping [ms]: avg 205.6 - min 204.4 (99.4%), max 207.4 (100.9%)
Web ping [ms]: avg 206.4 - min 204.5 (99.1%), max 231.0 (111.9%)
UK LON lon.speedtest.clouvider.net [F: 0]
DL [Mb/s]: avg 58.1 - min 57.5 (99.0%), max 58.7 (101.1%)
Ping [ms]: avg 192.8 - min 192.3 (99.7%), max 196.3 (101.8%)
Web ping [ms]: avg 193.5 - min 192.3 (99.4%), max 197.0 (101.8%)
NL AMS nl.mirrors.clouvider.net [F: 0]
DL [Mb/s]: avg 60.0 - min 59.4 (99.0%), max 61.0 (101.7%)
Ping [ms]: avg 186.9 - min 186.4 (99.8%), max 188.3 (100.8%)
Web ping [ms]: avg 187.2 - min 186.4 (99.6%), max 202.5 (108.2%)
DE FRA mirror.plusline.net [F: 0]
DL [Mb/s]: avg 57.1 - min 56.8 (99.5%), max 57.9 (101.3%)
Ping [ms]: avg 192.9 - min 0.0 (0.0%), max 197.0 (102.1%)
Web ping [ms]: avg 193.3 - min 0.0 (0.0%), max 208.1 (107.6%)
FR PAR mirror.in2p3.fr [F: 0]
DL [Mb/s]: avg 49.8 - min 48.6 (97.7%), max 50.7 (101.8%)
Ping [ms]: avg 223.4 - min 223.3 (100.0%), max 223.7 (100.1%)
Web ping [ms]: avg 494.0 - min 223.4 (45.2%), max 674.1 (136.5%)
CH GEN pkg.adfinis-on-exoscale.ch [F: 64]
DL [Mb/s]: avg 0.0 - min 0.0 (0.0%), max 0.0 (0.0%)
Ping [ms]: avg 197.9 - min 197.7 (99.9%), max 198.5 (100.3%)
Web ping [ms]: avg 197.9 - min 197.7 (99.9%), max 198.5 (100.3%)
IT MIL it1.mirror.vhosting-it.com [F: 0]
DL [Mb/s]: avg 65.4 - min 64.4 (98.6%), max 66.1 (101.1%)
Ping [ms]: avg 171.7 - min 171.6 (100.0%), max 171.8 (100.1%)
Web ping [ms]: avg 181.1 - min 171.6 (94.8%), max 334.6 (184.8%)
ES MAD mirror.es.stackscale.com [F: 64]
DL [Mb/s]: avg 0.0 - min 0.0 (0.0%), max 0.0 (0.0%)
Ping [ms]: avg 246.5 - min 246.3 (99.9%), max 247.1 (100.2%)
Web ping [ms]: avg 246.5 - min 246.3 (99.9%), max 247.1 (100.2%)
RO BUC mirrors.hosterion.ro [F: 0]
DL [Mb/s]: avg 47.0 - min 45.8 (97.6%), max 48.1 (102.4%)
Ping [ms]: avg 234.4 - min 232.6 (99.2%), max 239.0 (101.9%)
Web ping [ms]: avg 471.2 - min 462.7 (98.2%), max 533.5 (113.2%)
RU MOS mirror.yandex.ru [F: 0]
DL [Mb/s]: avg 54.3 - min 48.5 (89.4%), max 56.6 (104.3%)
Ping [ms]: avg 204.9 - min 204.0 (99.5%), max 246.4 (120.2%)
Web ping [ms]: avg 233.2 - min 204.0 (87.5%), max 416.0 (178.4%)
--- Asia / Oceania ---
RU SIB mirror.truenetwork.ru [F: 0]
DL [Mb/s]: avg 44.2 - min 42.3 (95.8%), max 45.9 (103.8%)
Ping [ms]: avg 248.6 - min 247.8 (99.7%), max 250.6 (100.8%)
Web ping [ms]: avg 252.1 - min 248.0 (98.4%), max 308.4 (122.3%)
IR SHI ir.almalinux.sindad.cloud [F: 12]
DL [Mb/s]: avg 34.3 - min 0.0 (0.0%), max 44.7 (130.6%)
Ping [ms]: avg 204.4 - min 0.0 (0.0%), max 277.6 (135.8%)
Web ping [ms]: avg 228.0 - min 0.0 (0.0%), max 1277.3 (560.3%)
IN PUJ in-mirror.garudalinux.org [F: 0]
DL [Mb/s]: avg 100.6 - min 88.9 (88.3%), max 110.4 (109.7%)
Ping [ms]: avg 106.5 - min 0.0 (0.0%), max 157.0 (147.4%)
Web ping [ms]: avg 111.6 - min 0.0 (0.0%), max 163.5 (146.6%)
SG SGP mirror.sg.gs [F: 0]
DL [Mb/s]: avg 333.5 - min 312.4 (93.7%), max 349.6 (104.8%)
Ping [ms]: avg 32.3 - min 32.0 (99.2%), max 34.5 (106.9%)
Web ping [ms]: avg 35.3 - min 32.8 (92.9%), max 39.6 (112.2%)
CN HKG mirrors.xtom.hk [F: 0]
DL [Mb/s]: avg 359.8 - min 350.0 (97.3%), max 364.9 (101.4%)
Ping [ms]: avg 30.3 - min 30.2 (99.6%), max 30.4 (100.3%)
Web ping [ms]: avg 31.5 - min 30.5 (96.9%), max 77.6 (246.7%)
CN BEJ mirrors.bfsu.edu.cn [F: 0]
DL [Mb/s]: avg 95.8 - min 58.1 (60.7%), max 104.7 (109.3%)
Ping [ms]: avg 105.0 - min 0.0 (0.0%), max 132.6 (126.2%)
Web ping [ms]: avg 108.1 - min 0.0 (0.0%), max 190.3 (176.0%)
JP OSA mirrors.xtom.jp [F: 0]
DL [Mb/s]: avg 67.2 - min 56.5 (84.1%), max 76.7 (114.2%)
Ping [ms]: avg 156.8 - min 146.1 (93.2%), max 172.4 (109.9%)
Web ping [ms]: avg 166.9 - min 146.5 (87.8%), max 193.7 (116.1%)
AU SYD mirror.internet.asn.au [F: 0]
DL [Mb/s]: avg 90.8 - min 86.0 (94.7%), max 92.3 (101.6%)
Ping [ms]: avg 124.9 - min 123.0 (98.5%), max 189.1 (151.4%)
Web ping [ms]: avg 125.6 - min 123.0 (98.0%), max 189.1 (150.6%)
--- Africa ---
ZA JOB mirror.datakeepers.co.za [F: 0]
DL [Mb/s]: avg 32.5 - min 31.9 (98.0%), max 33.0 (101.4%)
Ping [ms]: avg 348.6 - min 347.9 (99.8%), max 352.9 (101.2%)
Web ping [ms]: avg 368.2 - min 348.0 (94.5%), max 1389.8 (377.5%)
KE NAI mirror.liquidtelecom.com [F: 0]
DL [Mb/s]: avg 30.5 - min 29.3 (96.0%), max 31.7 (103.8%)
Ping [ms]: avg 354.5 - min 353.9 (99.8%), max 357.9 (100.9%)
Web ping [ms]: avg 584.8 - min 354.7 (60.7%), max 628.4 (107.5%)
--- America ---
US NYC nyc.mirrors.clouvider.net [F: 0]
DL [Mb/s]: avg 50.5 - min 49.4 (97.8%), max 51.6 (102.2%)
Ping [ms]: avg 220.7 - min 220.5 (99.9%), max 220.9 (100.1%)
Web ping [ms]: avg 223.0 - min 220.6 (98.9%), max 248.3 (111.3%)
US CHI ord.mirror.rackspace.com [F: 0]
DL [Mb/s]: avg 60.4 - min 55.1 (91.3%), max 65.1 (107.9%)
Ping [ms]: avg 185.0 - min 184.9 (99.9%), max 185.2 (100.1%)
Web ping [ms]: avg 193.1 - min 184.9 (95.7%), max 228.7 (118.4%)
US LAX mirror.alma.lax1.serverforge.org [F: 0]
DL [Mb/s]: avg 73.3 - min 68.2 (93.0%), max 79.6 (108.5%)
Ping [ms]: avg 158.2 - min 157.7 (99.7%), max 162.4 (102.6%)
Web ping [ms]: avg 159.2 - min 157.8 (99.1%), max 166.8 (104.8%)
US SJO mirrors.xtom.us [F: 0]
DL [Mb/s]: avg 60.3 - min 57.8 (95.8%), max 64.8 (107.4%)
Ping [ms]: avg 188.2 - min 188.1 (99.9%), max 188.4 (100.1%)
Web ping [ms]: avg 188.6 - min 188.1 (99.7%), max 199.0 (105.5%)
At first I was shocked when I saw the results for Europe, but of course we're looking at a HongKong VPS here.
As usual split into continents/regions.
Europe - meh, just meh, but of course other providers' VPS in HongKong will look similarly meh.
Btw, I - again - tried both the CH and the ES target from my PC and both work just fine. Must be something weird going on with routing I guess or maybe those targets filter out requests coming from a DC.
Asia / Oceania - half-cooked. Only one of three Iran targets with an acceptable result. OTOH both CN and all three Ozzyland target were reached fine and achieved quite decent results, the Ozzy ones even showed very similar results.
As for India I think this is the first time that I see webwerks beaten by garudalinux although both have similar routes, namely AS132337 (Axclusive) and Airtel. Weird.
Singapore with over 300 Mb/s seems to be quite good but HongKong just about 10% better - from HongKong and a decent DC, mind you - isn't something to write home about. China, Bejing with almost 100 Mb/s looks good though, well, as good as something routed by China Mobile for a very significant part of the route can look (In other words: look at their routes and you learn how not to do routing ...). Btw the secondary China target also achieved a quite decent although weaker result.
How one manages to achieve really decent 90 Mb/s to Ozzyland but only measly under 70 Mb/s to Japan from HKG escapes me, but oh well it is what it is.
Africa - meh, nuff said
America - quite meh as well. OK, it's quite a distance between HKG and say LAX but I for one am not at all impressed by those results.
Important note: those not really impressive results almost certainly are not @orangevps' fault! Nope, they chose a decent DC and, as far as I can tell, decent backend pipes which seem to very much prefer (just like me) twelve99 and besides a few cogent routes (not even bad surprisingly) and so do most intermediate segments ASs.
My still quite limited experience with Asia connectivity doesn't allow me to make bold statements, but after looking at the routes to all targets it seems that it's just really hard to get good (as is performance) connectivity there and the provider really did what's possible.
Comments
Now the Singapore VPS. As usual sysinfo, processor, and memory first
This one has a somewhat weaker E5v4 processor and the matching mediocre crypto results.
But at least all the commonly desired flags plus AVX2 are there.
Let's see how the disk performs ...
Yay, much better than the other one. Still no risk of breaking a performance record but perfectly fine for most work loads. No complaints at all.
Plus, let's be honest, most of us care most about connectivity with Asia VPSs anyway, so let's have a look how this box fares ...
Europe - rather meh again, just a bee stick better than the HKG one.
Asia / Oceania - Iran even way worse, but OK that may be due to the current situation. Re. India though my world is in balance again, albeit a bit brutally; webwerks is almost 10 times faster than garuda and easily blows the HKG result out of the water (+80%).
Singapore target is fast, duh, no surprise there, but the secondary SGP target also achieved a respectable result. HongKong achieved nearly the same result the VPS in HKG achieved, so really nice. CN however is about a third slower but still, at least in my books, any result solidly over 50 Mb/s in CN is a winner or at least really OK.
The result for Japan is about double at about 135 Mb/s; very decent. And Ozzyland also achives somewhat better results at over 100 Mb/s for all three targets; very nice as well. I really like what I see.
Africa - somewhat less meh but still meh.
America - clearly weaker than the HKG VPS.
TL;DR/verdict: A really decent VPS. Had I not seen and benchmarked the Leaseweb/@HostDZire Singapore VPS, I wouldn't hesitate to recommend this VPS as a very decent choice in Asia.
But I did. And so with that in mind I see a somewhat weak processor, similar not great but perfectly acceptable for most use cases disk performance and roughly similar connectivity modulo a few exceptions where one VPS is significantly faster or slower than the other one. And well noted, I'm not talking about Leasewebs stellar SGP performance because I guess nobody buys a VPS just to have outstanding connectivity within a single and small region.
As I said, normally I'd recommend this VPS, but I do not; not because it's somehow bad, it is not, but because the Leaseweb / HostDZire VPS has a significantly faster processor - and double the amount of cores - plus 50% more RAM for a similar price (in fact, even lower IIRC).
thank god i didnt go to bed because jsg benchmark just dropped
!!
Finally the US (Kansas City, Missouri) VPS. And it even has a Xeon Scalable Gold processor!
... but sadly the performance isn't there. But at least AES performance is in a decent territory
Maybe the disk is more attractive? Let's see.
Yep, it indeed is the best of the bunch. Not a speed daemon, by no means, but really decent. Nothing to complain about at all.
Now I'm looking forward to see really good and convincing connectivity within Murrica and maybe decent speed to Europe. Let's look at it
As usual split into continents/ regions
Europe - Yay, very decent results indeed!
Asia / Oceania - Also quite good results, overall not the best I've ever seen but really decent.
Africa - meh
America - YES! About 400 Mb/s to NYC, even a bit better to Chicago, and close to 250 Mb/s to the west coast. Me impressed, very nice results!
TL;DR/verdict: Yes, the results within Murrica are among the best I've seen so far and @orangevps smartly chose a good location not at either of the coasts which provides them with really decent and consistent connectivity within all of the country.
But alas, again there's also the Leaseweb (@HostDZire) WDC VPS. Yes, its connectivity within Murrica is less consistent (very high on the east coast but much less so to the west coast) but it has a way better processor, double the number of cores, 50% more memory and is cheaper and has overall somewhat better global connectivity.
So, my verdict is that this orangevps is a really decent product (keep in mind that I'm not hunting Ryzen results but prefer a well balanced product), but ...
I recommend it (only, sorry) if you care most or only about connectivity within the US. That's where this box shines and IMO even outperforms the Leaseweb WDC VPS.
If however you also care about global connectivity this VPS probably is not your best choice.
Hello,
Thank you for this honest review, we appreciate and review like this makes OrangeVPS keep growth more strong!