All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Romanian Shootout & HostSolutions Alternatives
With even the "working" and reachable HostSolutions VPS (except the Norway location) all but dead (see below) alternatives are obviously needed. While for most HS customers those do not necessarily need to be located in Romania I thought that it be interesting and worthwhile to first focus on romanian providers.For the sake of completeness and to make the situation clear I also tried to benchmark 2 of HS' current VPS.
Let's start with those HS VPS, first a somewhat larger one
Machine: amd64, Arch.: amd64, Model: Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz
OS, version: FreeBSD 12.2, Mem.: 3.988 GB
CPU - Cores: 2, Family/Model/Stepping: 6/79/1
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 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 hle avx2 smep bmi2 erms invpcid rtm rdseed
adx smap umip syscall nx pdpe1gb rdtscp lm lahf_lm lzcnt
ProcMem SC [MB/s]: avg 289.2 - min 285.2 (98.6 %), max 293.2 (101.4 %)
ProcMem MA [MB/s]: avg 501.8 - min 472.6 (94.2 %), max 531.0 (105.8 %)
ProcMem MB [MB/s]: avg 552.3 - min 527.2 (95.5 %), max 577.4 (104.5 %)
--- Disk - Buffered ---
Write seq. [MB/s]: avg 349.65 - min 287.69 (82.3%), max 411.61 (117.7%)
Write rnd. [MB/s]: avg 2882.64 - min 2822.69 (97.9%), max 2942.60 (102.1%)
Read seq. [MB/s]: avg 906.73 - min 901.23 (99.4%), max 912.23 (100.6%)
Read rnd. [MB/s]: avg 4896.63 - min 4792.44 (97.9%), max 5000.82 (102.1%)
--- Disk - Sync/Direct ---
Write seq. [MB/s]: avg 51.58 - min 50.80 (98.5%), max 52.36 (101.5%)
Write rnd. [MB/s]: avg 132.73 - min 132.53 (99.8%), max 132.93 (100.2%)
Read seq. [MB/s]: avg 556.92 - min 555.19 (99.7%), max 558.66 (100.3%)
Read rnd. [MB/s]: avg 195.90 - min 182.39 (93.1%), max 209.42 (106.9%)
--- Network ---
The processor performance is actually quite nice and all the nice E5-26xx v4 flags like AES, AVX2 etc. are available.
The NVMe is not so nice. While performance is very steady it's also underwhelming for a NVMe.
But the real bummer is the network. Almost all network tests completely failed; that's why that segment is empty above. Side note: one also really needs patience to establish an SSH connection ...
But hey, I have good will and so I scanned through the result sets by hand and found one with a result that is not a complete failure (cough, cough). Here you go
----- Network -----
[N] speedtest.lon02.softlayer.com UK LON:, P: 36.6 ms WP: 41.3 ms, DL: 73.05 Mb/s
[N] speedtest.c1.mel1.dediserve.com AU MEL:, P: 333.1 ms WP: 333.1 ms, DL: 0.00 Mb/s
[N] speedtest.che01.softlayer.com IN CHN:, P: 162.8 ms WP: 162.8 ms, DL: 0.00 Mb/s
[N] mirror.sg.leaseweb.net SG SGP:, P: 206.8 ms WP: 206.8 ms, DL: 0.00 Mb/s
[N] fra.lg.core-backbone.com DE FRA:, P: 55.9 ms WP: 55.9 ms, DL: 0.00 Mb/s
[N] speedtest.mil01.softlayer.com IT MIL:, P: 45.8 ms WP: 45.8 ms, DL: 0.00 Mb/s
[N] speedtest.par01.softlayer.com FR PAR:, P: 38.3 ms WP: 38.3 ms, DL: 0.00 Mb/s
[N] speedtest.hostkey.ru RU MOS:, P: 51.2 ms WP: 51.2 ms, DL: 0.00 Mb/s
[N] speedtest.sao01.softlayer.com BR SAO:, P: 241.8 ms WP: 241.8 ms, DL: 0.00 Mb/s
[N] speedtest.dal05.softlayer.com US DAL:, P: 142.3 ms WP: 142.3 ms, DL: 0.00 Mb/s
[N] speedtest.sjc01.softlayer.com US SJC:, P: 171.3 ms WP: 171.3 ms, DL: 0.00 Mb/s
[N] lax.download.datapacket.com US LAX:, P: 187.2 ms WP: 187.2 ms, DL: 0.00 Mb/s
[N] mirror.wdc1.us.leaseweb.net US WDC:, P: 134.6 ms WP: 134.6 ms, DL: 0.00 Mb/s
[N] nyc.download.datapacket.com US NYC:, P: 123.7 ms WP: 123.7 ms, DL: 0.00 Mb/s
[N] speedtest.tokyo2.linode.com JP TOK:, P: 0.0 ms WP: 0.0 ms, DL: 0.00 Mb/s
[N] 185.183.99.8 RO BUC:, P: 16.9 ms WP: 18.9 ms, DL: 24.84 Mb/s
[N] speedtest.ftp.otenet.gr GR UNK:, P: 0.0 ms WP: 0.0 ms, DL: 0.00 Mb/s
[N] 185.65.204.169 TR_UNK :, P: 48.9 ms WP: 48.9 ms, DL: 0.00 Mb/s
[N] speedtest.osl01.softlayer.com NO OSL:, P: 54.7 ms WP: 54.7 ms, DL: 0.00 Mb/s
[N] mirror.hk.leaseweb.net CN_HK :, P: 208.3 ms WP: 208.3 ms, DL: 0.00 Mb/s
"Only" 18 out of 20 tests failed ..
Funny side note: note that the download result for Bucharest, Romania is worse than that for London, UK. And how a ping packet could need almost 17 ms from Oradea to Bucharest remains a puzzle, I guess.
I'll spare us the results of the other, somewhat smaller, NVMe VPS which are pretty much the same.
NOTE: All VPSs from here on had >> 50 benchmark runs.
The next contender, /Dedicatserver.ro / @dedicatserver_ro. I benchmarked one of their €3/month offers here on LET (which was purchased normally).
Here are the results:
Machine: amd64, Arch.: amd64, Model: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
OS, version: FreeBSD 13.0, Mem.: 3.990 GB
CPU - Cores: 2, Family/Model/Stepping: 6/45/7
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 ssse3 cx16 pcid
sse4_1 sse4_2 x2apic popcnt tsc_deadline aes xsave osxsave avx
hypervisor
Ext. Flags: tsc_adjust syscall nx pdpe1gb rdtscp lm lahf_lm
ProcMem SC [MB/s]: avg 221.4 - min 203.4 (91.9 %), max 226.8 (102.4 %)
ProcMem MA [MB/s]: avg 421.6 - min 379.8 (90.1 %), max 452.1 (107.2 %)
ProcMem MB [MB/s]: avg 429.2 - min 390.4 (91.0 %), max 451.9 (105.3 %)
--- Disk - Buffered ---
Write seq. [MB/s]: avg 1438.46 - min 1208.51 (84.0%), max 1520.80 (105.7%)
Write rnd. [MB/s]: avg 4456.75 - min 3291.33 (73.9%), max 5149.43 (115.5%)
Read seq. [MB/s]: avg 4228.15 - min 2769.14 (65.5%), max 4948.94 (117.0%)
Read rnd. [MB/s]: avg 5872.72 - min 4485.34 (76.4%), max 7252.28 (123.5%)
--- Disk - Sync/Direct ---
Write seq. [MB/s]: avg 33.56 - min 29.47 (87.8%), max 37.26 (111.0%)
Write rnd. [MB/s]: avg 67.24 - min 57.31 (85.2%), max 74.30 (110.5%)
Read seq. [MB/s]: avg 2815.44 - min 2253.93 (80.1%), max 3138.27 (111.5%)
Read rnd. [MB/s]: avg 331.99 - min 293.29 (88.3%), max 377.59 (113.7%)
--- Network ---
US LAX lax.download.datapacket.com [F: 0]
DL [Mb/s]: avg 49.7 - min 34.6 (69.7%), max 63.5 (127.9%)
Ping [ms]: avg 189.8 - min 175.0 (92.2%), max 209.4 (110.3%)
Web ping [ms]: avg 192.7 - min 181.7 (94.3%), max 209.4 (108.7%)
NO OSL speedtest.osl01.softlayer.com [F: 0]
DL [Mb/s]: avg 153.6 - min 146.0 (95.1%), max 165.2 (107.5%)
Ping [ms]: avg 56.0 - min 53.8 (96.0%), max 70.4 (125.6%)
Web ping [ms]: avg 58.9 - min 54.0 (91.7%), max 219.3 (372.4%)
US SJC speedtest.sjc01.softlayer.com [F: 0]
DL [Mb/s]: avg 48.3 - min 33.8 (69.9%), max 52.6 (109.0%)
Ping [ms]: avg 173.8 - min 0.0 (0.0%), max 191.7 (110.3%)
Web ping [ms]: avg 179.4 - min 0.0 (0.0%), max 405.1 (225.8%)
AU MEL speedtest.c1.mel1.dediserve.com [F: 0]
DL [Mb/s]: avg 27.8 - min 25.5 (91.7%), max 29.7 (106.9%)
Ping [ms]: avg 323.3 - min 307.2 (95.0%), max 353.9 (109.4%)
Web ping [ms]: avg 324.4 - min 309.1 (95.3%), max 364.6 (112.4%)
JP TOK speedtest.tokyo2.linode.com [F: 0]
DL [Mb/s]: avg 40.8 - min 38.0 (92.9%), max 43.9 (107.6%)
Ping [ms]: avg 267.3 - min 264.8 (99.1%), max 284.9 (106.6%)
Web ping [ms]: avg 282.9 - min 266.7 (94.3%), max 366.3 (129.5%)
IT MIL speedtest.mil01.softlayer.com [F: 0]
DL [Mb/s]: avg 190.2 - min 103.2 (54.3%), max 225.5 (118.6%)
Ping [ms]: avg 43.8 - min 0.0 (0.0%), max 65.3 (149.2%)
Web ping [ms]: avg 75.5 - min 0.0 (0.0%), max 1415.5 (1874.4%)
TR_UNK 185.65.204.169 [F: 0]
DL [Mb/s]: avg 87.3 - min 86.7 (99.3%), max 87.5 (100.3%)
Ping [ms]: avg 43.5 - min 43.2 (99.3%), max 45.2 (103.9%)
Web ping [ms]: avg 43.8 - min 43.4 (99.1%), max 49.1 (112.1%)
FR PAR speedtest.par01.softlayer.com [F: 6]
DL [Mb/s]: avg 160.6 - min 0.0 (0.0%), max 196.2 (122.2%)
Ping [ms]: avg 47.5 - min 45.1 (94.9%), max 63.6 (133.8%)
Web ping [ms]: avg 47.9 - min 45.2 (94.4%), max 68.5 (143.1%)
SG SGP mirror.sg.leaseweb.net [F: 0]
DL [Mb/s]: avg 62.9 - min 59.2 (94.1%), max 65.1 (103.6%)
Ping [ms]: avg 177.7 - min 0.0 (0.0%), max 231.0 (130.0%)
Web ping [ms]: avg 180.7 - min 0.0 (0.0%), max 231.0 (127.8%)
BR SAO speedtest.sao01.softlayer.com [F: 0]
DL [Mb/s]: avg 38.3 - min 34.9 (91.1%), max 43.7 (114.1%)
Ping [ms]: avg 216.2 - min 0.0 (0.0%), max 235.4 (108.9%)
Web ping [ms]: avg 234.3 - min 0.0 (0.0%), max 998.1 (425.9%)
IN CHN speedtest.che01.softlayer.com [F: 4]
DL [Mb/s]: avg 46.7 - min 0.0 (0.0%), max 55.2 (118.1%)
Ping [ms]: avg 166.0 - min 0.0 (0.0%), max 185.1 (111.5%)
Web ping [ms]: avg 187.2 - min 0.0 (0.0%), max 1262.4 (674.5%)
GR UNK speedtest.ftp.otenet.gr [F: 30]
DL [Mb/s]: avg 314.4 - min 0.0 (0.0%), max 661.8 (210.5%)
Ping [ms]: avg 9.1 - min 0.0 (0.0%), max 20.3 (223.9%)
Web ping [ms]: avg 30.6 - min 0.0 (0.0%), max 685.1 (2242.4%)
US WDC mirror.wdc1.us.leaseweb.net [F: 0]
DL [Mb/s]: avg 99.4 - min 89.0 (89.4%), max 102.4 (103.0%)
Ping [ms]: avg 113.0 - min 112.6 (99.6%), max 115.5 (102.2%)
Web ping [ms]: avg 114.8 - min 112.7 (98.2%), max 128.2 (111.7%)
RU MOS speedtest.hostkey.ru [F: 0]
DL [Mb/s]: avg 176.1 - min 151.0 (85.8%), max 195.2 (110.9%)
Ping [ms]: avg 61.9 - min 0.0 (0.0%), max 76.7 (123.9%)
Web ping [ms]: avg 64.6 - min 0.0 (0.0%), max 79.3 (122.8%)
US DAL speedtest.dal05.softlayer.com [F: 2]
DL [Mb/s]: avg 55.2 - min 0.0 (0.0%), max 63.4 (114.9%)
Ping [ms]: avg 142.8 - min 0.0 (0.0%), max 163.6 (114.6%)
Web ping [ms]: avg 160.9 - min 0.0 (0.0%), max 990.8 (615.9%)
UK LON speedtest.lon02.softlayer.com [F: 0]
DL [Mb/s]: avg 215.7 - min 122.4 (56.8%), max 239.6 (111.1%)
Ping [ms]: avg 40.2 - min 15.6 (38.8%), max 54.6 (135.7%)
Web ping [ms]: avg 77.8 - min 38.0 (48.9%), max 1467.0 (1886.0%)
US NYC nyc.download.datapacket.com [F: 0]
DL [Mb/s]: avg 92.3 - min 81.6 (88.5%), max 95.5 (103.4%)
Ping [ms]: avg 121.8 - min 119.6 (98.2%), max 135.4 (111.2%)
Web ping [ms]: avg 122.7 - min 119.6 (97.4%), max 146.1 (119.0%)
RO BUC 185.183.99.8 [F: 0]
DL [Mb/s]: avg 1016.7 - min 991.7 (97.5%), max 1040.7 (102.4%)
Ping [ms]: avg 0.7 - min 0.5 (69.5%), max 1.9 (264.1%)
Web ping [ms]: avg 1.1 - min 0.7 (61.7%), max 10.2 (899.6%)
CN_HK mirror.hk.leaseweb.net [F: 0]
DL [Mb/s]: avg 52.9 - min 48.8 (92.2%), max 56.0 (105.9%)
Ping [ms]: avg 217.6 - min 0.0 (0.0%), max 249.6 (114.7%)
Web ping [ms]: avg 219.3 - min 0.0 (0.0%), max 249.6 (113.8%)
DE FRA fra.lg.core-backbone.com [F: 0]
DL [Mb/s]: avg 395.7 - min 228.4 (57.7%), max 472.3 (119.3%)
Ping [ms]: avg 30.3 - min 0.0 (0.0%), max 48.0 (158.3%)
Web ping [ms]: avg 31.5 - min 0.0 (0.0%), max 77.8 (246.7%)
The processor seems to be an old E5-26xx v2 but the results have low spread and the multi-core result is almost double of the single core result which is nice. As an added plus it seems that the nodes aren't too crowded or oversold.
AES, AVX(1), and even hypervisor flags are available. Nice.
As for the disks the results are somewhat mixed and in part even disappointing. The reason seems to be that some kind of "Cloud disks" are used (along with a large cache). So the buffered NVMe results are really nice but the direct/sync results are poor and largely about at the level of a crappy SSD.
Short version: OK for normal files but to be avoided for DB servers.
And it gets even worse for the SSD (the product comes with 30 GB NVME + 50 GB SSD). Its buffered speed is halfway decent except for sequential writes but severely drops down to the level of a crappy HDD in direct/sync mode (e.g. write seq < 10 MB/s). Sorry, but that's the kind of performance I expect from HostSolutions, not from a decent provider.
The network though is really decent, albeit quite Europe centric (which is not necessarily a bad thing, depending on the customers needs). Almost all european targets (except Turkey which however still is decent) are above 100 Mb/s and some important ones like FRA, DE and LON,UK show excellent results - for a €3/mo VPS, mind you. The connectivity is officially limited to 100 Mb/s but as the results show, one can reach more (with a reasonable albeit unknown traffic limit I guess).
The weak points I see are in their infrastructure and panel. For one there is no remote console available in the panel, which I consider a major, even a critical minus. Plus, there is just a limited selection of (not always up to date) OSs to choose from and those get installed in a standardized way using up the full NVMe.
Plus, of course, there is the big Mega-Minus' of them logging into customer VPS when they feel like it that is, when they have a suspicion that a customer might have acted against their TOS/AUP.
I really wish they would clarify and change their position towards a more professional and privacy respecting attitude, because despite the not so great disk performance I would find their offer attractive all in all. 2 vCores, 4 GB mem. and 30 GB NVMe + 50 GB disk for €3/mo is not a bad deal at all.
Comments
The next candidate, Virtono / @virtono, is a provider I've come to like as a super-cheap provider with reasonably decent products at insane prices. The VPS I benchmarked is a (KVM based, of course) "Cloud VPS S" for € 23.70/year which is a bit less than €2/month (if payed annually). There is also an even cheaper a bit smaller VPS available for just €14.95/year.
Here are the results
The processor performance is a bit lower than others but still quite decent and the AES flag is available. And as the numbers show performance is steady with very low spread.
The 25 GB SSD seems to be less impressive than the now trendy NVMes but in direct/sync mode it's actually faster than some of the NVMes seen in this shootout. It's certainly no speed demon but I'd feel better running a DB (e.g. a dynamic web site) on this SSD than on the dedicatserver NVMe.
As for the network (traffic volume 2TB/mo) it's decent, not quite as good as dedicatserver but still the most important european targets are above 100 Mb/s. If really good low latency connectivity is important for your use case you might be better served alsewhere though.
As I happen to have their smaller 'XS' VPS since about 2 years I can report about my experience. In short, I've been really pleased. The VPS is a wee bit more "expensive" than the cheapest one can find (for less than $10/yr) but I wouldn't want to exchange. The Virtono support has always been reasonably fast (in the tens of minutes to a few hours range), friendly, and competent. Plus and more importantly IIRC that VPS which serves a one of my main name servers has never let me down (and if it did, it was fixed quickly so I didn't notice).
The panel is standard and offers a remote console and while (afaik) one can't upload custom ISOs I do remember that when I asked support they (quickly) took care of it and provided me with the ISO I wanted. Another point I really like that I'm not limited to template installations but I can configure everything to my liking. Not a big thing, you might think, but my experience shows that very low price and flexibility usually don't go hand in hand.
TL;DR A rather small but decent and reliable VPS with good support and a very attractive price (if payed annually).
Part 3
Now to one of the two "killer offers" (as I look at them), the HostHatch / @hosthatch offer. This is again a VPS that I purchased myself plus I payed 2 years in advance so I got double the RAM (8 GB), double the disk (40 GB), +15 TB traffic volume/mo.
For $30/yr! That's insane!
Here are the results
Yes it's just a 26xx v2 (like the others) with AES and hypervisor flags, but with decent RAM speed albeit with a (tiny) bit more spread. Nice!
The NVMe, while not at the high end, is really decent and consistent. If the annual price was $50+ I might finally have found something to complain about (being very picky) but for $30 per year all you'll get to hear from me is "thank you, Hosthatch, I'm very happy".
As for the network I'm stunned. Of bloody course my (really cheap, mind you!) hosthatch VPS in Oslo does over 100 Mb/s across almost all of Europe, and of course it does multiple 100 Mb/s to the most important targets, but: it also does over 100 Mb /s across the Atlantic to the US east coast! For $30/year!!
IMO this HostHatch offer (without any evil intention by them) is the HostSolutions killer. Against dedicatserver cociu could argue privacy, against Virtono he might try to bet on them not being aggressive in marketing - but against this HostHatch offer I do not see anything HostSolutions could put against with even a vague chance.
TL;DR If you want a nice combination of low price, good quality and performance, and decent support, look no further. These HostHatch offers are the ones to go for.
Finally the offer I consider to be the other killer. Virmach's / @VirMach current offer. I purchased one, in part because I've heard a lot about them, much of it positive, and wanted to experience their service myself. And, you guessed right, of bloody course I went right away and benchmarked my new (NL) VPS. Here is what I got for my $9/annually: 1 vCore, 512 MB mem., 10 GB SSD, 1 TB traffic (but the panel says I have 2 TB).
Here are the results
The processor seems to be a 26xx v2 but there is something strange: it passes the hypervisor flag through but not the AES flag. Weird. Maybe a config mistake?
Whatever, the performance is OK and decent, similar to other VPS with some of the better v2 26xx and decent speed memory. Me happy, no complaints (but maybe not the ideal VPS for a VPN).
The disk is really nice for a SSD; we've seen considerably worse NVMes in this shootout. Decent, me happy.
The network is OK too. A bit less performant than e.g. HostHatch but still nice. I'll certainly not complain about 500+ Mb/s to the major european targets, although even that is an indicator of some inconsistency. All in all Virmach's super cheap VPS is no disappointment, quite the contrary.
And it's the other potential HostSolutions killer I see here. While HostHatch leaves zero room in the low to mid range, Virmach's low end offer will attract those who are looking for reasonable quality at a very low end price.
About the only thing one could complain about is Virmach's support. It's professional, helpful, and friendly, but it seems to be (occasionally?) snail slow. Once they worked on my ticket it went smooth - but I had to wait almost a full day (ca. 20 hrs). But of course corners must be cut somewhere at that price range.
All in all, I'm happy and my (albeit short) experience with Virmach so far was very positive.
P.S. sorry forgot it: and this baby comes with a free Ryzen upgrade path! Yay!
Bye, bye HostSolutions. You should have communicated. Now we have way more attractive options to look at.
Seems like he still got everyone's prepaid service money and doubled up account credits - so who really wins here?
One might think that he is the winner because he gets to keep the money.
But I think that's short sighted for two reasons: (a) if he indeed just keeps the money, that would constitute fraud and he'd highly likely lose more than that money, and (b) if his situation should be bad enough to even consider just keeping the money, he'd lose it anyway to pay debts to companies and people who will fight (e.g. because they'd loose 4 or 5 figure sums which to fight for makes financial sense), lawyer and court costs, etc.
But frankly, I think in this thread we should talk about the benchmarks/reviews and the tasty alternatives to the a__hole in Oradea.
Awesome benchmarks. And I see you like FreeBSD too.
Thank you! But re. FreeBSD the reason actually is twofold, (a) FreeBSD gives more honest results, and (b) consistency. I want people to be able to compare benchmark results.
Personally I'm still a linux user and I wouldn't consider to switch to [any]BSD for my workstation. For servers though I might consider it.
FreeBSD seems not playing well with the common virtio driver found in every vps, causing worse network performance than linux. Would you boot your boxes into a linux live cd and do those tests again?
If you feel it's needed just do it yourself. My benchmark software if freely available and afaik all the VPSs I tested can still be purchased.
@dedicatserver_ro is also a pain in assholes who will see in your garden and snoop on your server's filesystem.
Worth mentioning for those who just tuned in..
He is not an official paid benchmark person of LET. He did this review by passion, out of his own free time. Asking him to redo it all over again (without even using the word "please" to ask politely) is like acting bossy on other people's energy, money, and time.
Feel free to rent such servers yourself, test them, bench them, and come up with your own constructive review, showing that Linux performs better compared with *BSD. This way we would all learn something new regarding "every vps", and it would start a very interesting debate. Until then, please, for the love of humanity, do not put others to test your imaginative theories.
@dahartigan 👏 👍
@jsg
You can't test apples with pears for example
1.you compare internal direct access SSD/NVMe-cached --> with RDMA network access SSD/NVMe-non cached but you tell cached just because it is the command, internal disk with external storage
Internal NVMe RAID 1(the worst for NVMe you lose 50% from speed ) with RAID 0 ( 2 x NVMe win 20% ) and external cluster storage ( you lose up to 20% from speed but your data are safe )
differently WM configuration CPU the higher the frequency the better the results MEM 2GB/4GB/8GB =>different cache in test
OVZ,QCOW and QCOW2
Some have a super suport to the others no word about
Next time start by describing the conditions under which you made such a test, and if you know the configurations behind it, what supervisor ... storage type .... etc
NO CACHE - RDMA --> remote direct memory access (RDMA) over an Ethernet network
virtio gives more honest results with centos and also depends on the template
Romanian Shootout & HostSolutions Alternatives
Virtual Machine Solutions LLC - @VirMach - not Romanian
HostHatch LLC - @HostHatch - not Romanian
Romanian company not tested providern on LET:
.....and more
@dedicatserver_ro is desperate to be the new cociu.
Uh you have a mistake apparently, it should be Virtono instead which is Romanian
@dahartigan 👏 👍
we can´t because of this:
Project: “Creation of a cluster of computing systems intended for operation in“ Cloud Computing ”” -SMIS 109478
For detailed information on the other operational programs co-financed by the European Union www.fonduri-ue.ro
not @Calin ?
I think that @Calin is now LowEndCociu.
Both are gunning for the title, but @dedicatserver_ro is extremely desperate.
Calin has a wooden rack, servers at home in the basement and its RAM is SWAP on cheap Kingston SSD.
This is not to lowend. He needs openvz, nulled whmcs and 100Mbps residiential. He has 1.5Gbps residential. Thought, it could be just 4g internet.
what is the network blend?
there is also @FlorinMarian with hazi.ro for romania hosting providers
It's mixed, not blended. He has telia. drama.gg has more info on him.
He had openvz, he had nulled whmcs and the internet speed according to benchmarks is 50 mb/s. He even sold vps that were on vps from contabo (double SWAP ram on disks?)
Benchmark: https://www.lowendtalk.com/discussion/comment/3262679/#Comment_3262679
In case you're affected see how you can cross romanian border.
https://youtube.com/watch?v=EEPCH_WxFL4
That's barrel bottom. F*ck that.
it's not swap , but yes, in the past I had openvz, and I was a reseller each has a startup.
It's simple. I measure what the customer gets in terms of performance. Of course there's a thousand different configurations behind the diverse VPS products of the diverse providers - but at the end of the day we customers care about the performance we get. And that's what I measure.
Yes, including support in a review is a complicated thing, so I mention it only if and when I have a reasonable basis in personal experience.
No, I clearly show the average as well as the best and the worst results.
The conditions are always the same and what most customers do. Install the OS (always almost the same), configure all basics (always almost the same), and then benchmark without any other major service (e.g. web server) running.
True, not romanian - but alternatives and with currently active offers.
Oy vey, I hope I can be forgiven that I did not purchase a VPS from every romanian provider. No evil reasons whatsoever and I'll be happy to test (almost) any provider who gives access to a free test VPS (either for 3 days or for 2 or 3 weeks, as they please).
As for yourself, you have nothing to complain about. I purchased your currently active VPS offer, paid for a year in advance and gave you a really fair shot, exactly the same chance all providers I test get.
How about some real world load and not synthetic tests? For example how provider reacts on hammering cpu on 85-95% constantly? How about constant use of bandwidth? They advertise 1Gbps? Push that to the max as long as you can and observe.
To rely on synthetics is not a wise decision. Torture that server, make cpu, disk and internet pipe suffer. Than you will see how budget provider differs from prem ones.
Funny enough, when I had vps with Borta, I hammered it to the max. And no complaints!