Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


dedicat_server.ro benchmark & review
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

dedicat_server.ro benchmark & review

jsgjsg Member, Resident Benchmarker
edited November 2019 in Reviews

First, kudos to the dedicate_server.ro people because they provided me with access to both their small KVM and their small server products. Many probably know them from their, uhm, critical remarks in @cociu's / HostSolutions threads. But, and that's laudable, they didn't just leave it at declaring their superiority but let me benchmark and review their goods.

Front up a declaration of my position in that: I'm on neither side. For me they are both providers with servers in Romania and I have no particular like or dislike of either. I have (since some days) one of HostSolutions storage VPS but that's only due to it's good value for a very low price. I also have a VPS with Virtono, another player on the romanian market. In short, you can be sure that I have no dog in that fight and am guided only by facts and numbers.

Now, let's get started, dedi first. System Info:

Machine: amd64, Arch.: amd64, Model: Intel(R) Xeon(R) CPU           L5630  @ 2.13GHz
OS, version: FreeBSD 12.0, Mem.: 15.974 GB
CPU - Cores: 4, Family/Model/Stepping: 6/44/2
Cache: 32K/32K L1d/L1i, 256K L2, 12M L3
Std. Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
          pse36 cflsh ds acpi mmx fxsr sse sse2 ss htt tm pbe sse3 pclmulqdq
          dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca
          sse4_1 sse4_2 popcnt aes
Ext. Flags: syscall nx pdpe1gb rdtscp lm lahf_lm

So, it's an old Westmere Xeon which is not a bad (or uncommon) choice for low end dedis. For users it's OK because it has all the usualy bells and whistles like aes, nx, VT, etc, and for providers it's attractive due to its low power requirements. Unless you have solid reasons to need, say an E-26xx v3, this processor will serve you well for a low price (also in the long run, because guess who's going to pay - or save - on the difference between a 50W and a 120W processor ...).
Btw, note that my test dedi has 16 GB memory, while normally it has 8 GB.

Processor & memory:

                        Avg   Min   Max Dmin   Dmax   D-Span
Single Core:   193.43   193.37   193.48   0.03   0.02   0.06
Multi Core:     806.17   772.62   826.87   4.16   2.57   6.73

Not bad. And as usual when you stress hyperthreading the result is much closer to hw cores than to ht cores. But that it does nicely. The multithreading results correspond indeed to 4 x single core performance (with 8 threads running).

Disk:

Seq.Wr.:    421.73  406.13  433.18  3.7     2.71     6.41
Rnd.Wr: 10.85   10.27   11.28   5.34    3.97    9.31
Seq.Rd: 1391    106.11  3679    92.37   164.4   256.78
Rnd.Rd: 2839    968.59  2949    65.88   3.87   69.76

The disk seems to be solid SSD. Not much to say other than it shows typical SSD results. Nothing to complain about but neither something to write home about.

Network:

GR_UNK: 109 3.1 127.3   97.16   16.79   113.94
US_SJC: 25.72   2.6 31.1    89.89   20.9    110.79
OK_LON: 107.28  6.8 147.1   93.66   37.12   130.78
DE_FRA: 153.12  7.7 205.7   94.97   34.34   129.31
BR_SAO: 19  0   24.8    100 30.49   130.49
FR_PAR: 107.5   6.8 157.8   93.67   46.78   140.46
US_WDC: 33  0   47  100 42.42   142.42
RO_BUC: 847.3   671.4   928.5   20.76   9.58    30.34
AU_MEL: 17.31   3.5 21.5    79.79   24.17   103.96
IT_MIL: 118.92  6   154.5   94.95   29.92   124.87
SG_SGP: 12.49   0.63    29  94.98   132.16  227.14
US_DAL: 30.51   3.3 37.6    89.18   23.24   112.42
RU_MOS: 98.93   89.1    107.9   9.94    9.07    19
JP_TOK: 9.99    0   21.1    100 111.14  211.14
NO_OSL: 66.4    6.6 88.5    90.06   33.29   123.35
IN_CHN: 25.68   7.9 34.5    69.24   34.34   103.58

First, the good news: for a dedi with an official limit of 100 Mb/s it's nice to see some prominent european targets above 100 Mb/s. The RO_BUC result, probably within the DC or anyway very close by, shows that dedicat_server really has good network hardware and config.
The bad news is that while the average results are within what could be called "normal", the Span is really high, meaning that many locations are almost like playing the lottery. You might have a fast connection but you might also have a poor connection.
I'll refrain from writing what I suppose to be the culprit but I'll clearly say that those results are no basis to brag about ones networking prowess, sorry.


Now the KVM VPS. First the system info:

Machine: amd64, Arch.: amd64, Model: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
OS, version: FreeBSD 12.0, Mem.: 1.985 GB
CPU - Cores: 1, Family/Model/Stepping: 6/45/7
Cache: 32K/32K L1d/L1i, 2M L2, ? 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

Well, a Xeon E-26xx, no need to say much about that. Nice processor, even the older generations. 2 MB L2 cache is nice and all the usual flags are passed through.

Processor & memory:

Single core:    223,68  195,36  228,94  12,66   2,35    15,01
Multi core: 254,54  211,96  266,45  16,73   4,68    21,41

First, as side note. Remember what I said above about the L5630? That old Westmere achieves about 85% of the E-2670 performance! Just mentioning it because many people seem to think that there is no life before E-26xx v3 - that's obviously wrong, especially when you care for bang for the buck.
Not much else to say other than the span suggests that that node is tightly packed. Not yet on the bad side but certainly tighly packed.

Disk:

Seq.Wr.:    204,95  74,54   265,76  63,63   29,67   93,3
Rnd.Wr: 14,69   13,41   15,66   8,72    6,59    15,31
Seq.Rd: 395,48  250,53  499,95  36,65   26,41   63,07
Rnd.Rd: 666,1   290,29  801,87  56,42   20,38   76,8

Meh, not so nice. Looks like a SAS or SATA spindles array. If it's SSD then it's a really poor one. That's not to say it's plain sh_tty, but it's certainly pulling down the verdict; That machine has the potential for more and the disks are spoiling it.

Network - Same as above.

Summary: Not recommended. Not as in "that's sh_tty", it's not, but as in "you'll get much better elsewhere for the money". As for the dedi, the system isn't bad, not at all, but it's simply way too expensive.

Important side note: I don't know how good or bad their support is because my interaction was quite limited; they reacted quickly however. THAT could be the factor to survive next to HostSolutions whose support is ridiculously poor. Probably they should rethink their pricing, too, or at least have more promos plus they should show a much more constructive attitude here at LET. And I'd suggest that dedicate_server seriously thinks about that advice because IF (a very big if) HostSolutions did improve their support drastically, I wouldn't see much of a future for dedicat_server, certainly not in the low and mid range segment.

(P.S. As usual, please forgive my sh_tty formatting)

«13

Comments

  • Can you explain the random writes and why those numbers make a "solid SSD"? Maybe it's a formatting problem, but looks like 10.85MB/s. It would be better if you included headers with units, that's just basic benchmark and review stuff.

  • jsgjsg Member, Resident Benchmarker

    @TimboJones said:
    Can you explain the random writes and why those numbers make a "solid SSD"? Maybe it's a formatting problem, but looks like 10.85MB/s. It would be better if you included headers with units, that's just basic benchmark and review stuff.

    • I already did explain that multiple times (and yes, the disk values are in MB/s)
    • I do understand that my way of doing it feels strange to some and I'm planning to develop a new version (sometimes, when I find time) with something that matches the usual outputs more.
    • but: My point was not to create yet another bla_bench providing nice but meaningless numbers.
      My point was to really check and benchmark systems, sometimes deliberately "breaking through" some common tricks like caches. I'm interested what a potential customer really gets and not in some impressive but largely meaningless numbers.

    • Yes, I confess it, my presentation is ugly and lacking, but hey, I'm a developer (not web!) and if it's not easy to format something nicely then I don't do it, simple as that.
      On the other hand I put a lot more effort in proper benchmarking and processing the result sets, as well as in interpreting them. And I'm investing typically quite some hours unlike someone who simply runs bla-test (twice if you are lucky) and is done. So, be a bit generous with me.

    Thanked by 1poisson
  • This was an interesting read - a rollercoaster of technical praxis that had me on the edge of my seat from the very beginning.

    While covering the usual formulaic elements of a lowend potboiler, the iterated subplots and true-to-life character development made for a refreshing take on a well-worn theme.

    Set in modern day Romania, we find the hero protagonist beset with conflicting desires and afflictions as they venture into "the net" - and yet they manage not only to prevail, but to eventually triumph over the confusion - guided through the haze as if lead by the nose by their ever-vigilant shepherds Ethos, Pathos, and Logos.

    I came away from the review feeling both humbled and edified by the uplifting yet cautionary tale. While some of the content may not be particularly palatable for children of any age, there was plenty of good food for thought for curmudgeons and shitposters alike.

  • Totally appreciating your effort and the writeup, but what I don't understand is what is the exact purpose of your reviews in terms of value for other users on this website?

    First of all people have to give you reasonable doubt that this is not a paid review, which I think you earned at least in my case with all the previous reviews you have written up that weren't always favorable for the provider and overall looked honest. Doesn't mean I am right, but I personally believe you try to be honest and have some integrity with your reviews and it's nice to see that people like you still seem to exist.

    But that out of the way, in terms of making a buying decision the review has absolutely zero value for a potential customer.

    The reviews I have seen from you are always done on instances that are specifically provided to you by the provider. You also don't provide benchmarks for a reference system that is regularly provided and the provider doesn't know belongs to you.

    Without a reference system that has been provided like any other customers, you have absolutely no idea if the provider manipulates your testing environment in any form or shape f.e. by putting you on a node that is specifically selected or manipulated with artificial load and other shenanigans.

    So you get good values out of your tests that don't necessarily reflect a normal node the average customer is put on while not being completely over the top as well. Like how do you want to provide a review that represents an authentic user experience without being a normal paying user?

    If you give a provider the opportunity to provide you a system for testing there is always room for manipulation.

    Even if that is not the case in the VPS market you are always reliant on how well your neighbors behave and that changes over time with old customers dropping off and new customers being boarded onto the same node as you. Projects grow or shrink, ppl abuse Resources etc.

    So you are basically providing a tiny snapshot of the performance of a VPS at the time it was provided to you by the provider on that specific node it's provisioned on at the specific time of your testing.

    That doesn't mean I get the same experience and frankly I tried a view of the providers you reviewed and got wildly different experiences with some of them.

    Lastly talking about authenticity, how can I make sure you actually did test those products as someone who is interested in maybe renting from this specific provider you reviewed?

    I can just copy & paste stuff like that and change the values

    jsg said: Seq.Wr.: 421.73 406.13 433.18 3.7 2.71 6.41

    Rnd.Wr: 10.85 10.27 11.28 5.34 3.97 9.31
    Seq.Rd: 1391 106.11 3679 92.37 164.4 256.78
    Rnd.Rd: 2839 968.59 2949 65.88 3.87 69.76

    So basically, if I found your threads and wanted to educate myself about the performance of a provider or specific value of an offer

    I can't

    a) verify that you didn't get paid by the provider for this, especially since you did not spend any money yourself trying to get a "neutral" testing environment.

    I can't

    b)verify the authenticity of your data because you don't provide me with proof you actually had access to those products and did test on them. Something like a YT Video or anything
    where you show that the system and IP belongs to the provider in charge and the actual testing procedures would give your data a lot more credibility

    and lastly

    You are providing a snapshot of the performance of a product for a very small timeframe mostly. So it's not even guaranteed I get anything close / better / worse than what you are showing if I order the same product two weeks later but you never mention this in any of your reviews.

    Like I said I respect all the effort you put into those reviews but I think you should overthink your process and also provide some kind of disclaimer that this is just your experience over a very limited timeframe for specific product a/b and that it's not usable or reliable to base a buying decision or any form of judgment about a provider or their products performance on. "Your mileage may vary".

    F.e.

    jsg said: THAT could be the factor to survive next to HostSolutions whose support is ridiculously poor.

    My last ticket with HS was 26/08/2019 (04:22) and it was answered at 26/08/2019 (07:13) and problem solved by 26/08/2019 (07:24). Besides that I had three months of perfectly fine performance with them ( besides the DDOS downtimes ).

    I still didn't renew because of all the problems they had but I really can't complain much about their service.

    So the question still is what exactly is the purpose of those reviews and what kind of value do they provide for me who's potentially interested in trying out new providers and offers for small / personal / side projects?

  • dont know why they have bad blood with HS (not that HS is impeccable) but they should have focused all that energy on improving their services and website

  • Who is considered to be 'premium' provider in Romania?

  • tgltgl Member
    edited November 2019

    @LTniger said:
    Who is considered to be 'premium' provider in Romania?

    I doubt there are any premium ones for virtual private servers, or at least this was the impression the last time I checked, mainly because the of the infrastructure (very few data centers), but that will change in the near future.

    You can get 'premium' support, and 'premium' hardware, the peering is good but the providers are limited, 2N redundant and cooling (Tier 4) are not present in many places.

    I saw some things I liked about @intovps , but I did not try them, still they seem to offer some products that are trendy now (e.g. openstack)

    Thanked by 1intovps
  • @cybertech said:
    dont know why they have bad blood with HS (not that HS is impeccable) but they should have focused all that energy on improving their services and website

    because @cociu got in touch with @dedicatserver_ro 's sister

  • @EAgency said: The reviews I have seen from you are always done on instances that are specifically provided to you by the provider. You also don't provide benchmarks for a reference system that is regularly provided and the provider doesn't know belongs to you.

    Without a reference system that has been provided like any other customers, you have absolutely no idea if the provider manipulates your testing environment in any form or shape f.e. by putting you on a node that is specifically selected or manipulated with artificial load and other shenanigans.

    I made this point some months ago in connection with a review of his, and I was disciplined for daring to say this

    Thanked by 1skorous
  • dedicatserver_rodedicatserver_ro Member, Host Rep
    edited November 2019

    First on all - Thanks! @jsg

    jsg said: First, the good news: for a dedi with an official limit of 100 Mb/s it's nice to see some prominent european targets above 100 Mb/s.

    • each dedicated server have 1Gbps port open, the client can use the full speed on short time period ( 10-20 min ).

    jsg said: Meh, not so nice. Looks like a SAS or SATA spindles array. If it's SSD then it's a really poor one. That's not to say it's plain sh_tty, but it's certainly pulling down the verdict; That machine has the potential for more and the disks are spoiling it

    • we use one "Distributed NVME-SSD Tiring Storage" ( see the picture, it's something like Ceph ), you don´t W/R direct on one SSD/SSD RAID
    • if you disable the NVME-SSD Tiering(caching) you will W/R on the FS ( 3/7 - 3 x Metadata servers + 7 Chunk Servers)

    LTniger said: Who is considered to be 'premium' provider in Romania?

    If we are a premium provider ...can´t tell that, but :

    • we are transparent and honest provider
    • about me
    • our servers are colocated in Voxility and M247 - both DC are Tier4
    • we use multiple internet conections and are RIPE members with own IPs -AS3164
    • we have a full mesh SDN network with enterprise hardware and our own DDOS protection ( see down )
    • all our hardware ( servers,switches, firewall...ddos) have dual redundant power-supply,IPMI-ILO-AMM access, enterprise SSD/HDD/NVMe, DDR ECC...etc.
    • ticket answer in max 15 min.

    • DDOS

  • angstromangstrom Moderator
    edited November 2019

    @EAgency said: Like I said I respect all the effort you put into those reviews but I think you should overthink your process and also provide some kind of disclaimer that this is just your experience over a very limited timeframe for specific product a/b and that it's not usable or reliable to base a buying decision or any form of judgment about a provider or their products performance on. "Your mileage may vary".

    I also appreciate the effort that @jsg makes in his reviews (and have often thanked him for this), but I also often find the tone and the buying recommendation (yes/no) a bit too authoritative (also for the reasons that you mentioned)

  • cociucociu Member
    edited November 2019

    I have only one question in all this @dedicatserver_ro what you think you will win if you blame another provider ? what you win if you post here good pictures (like always including in my offer threads) , put your ass resolve your network and start to sale , like you i have missing so try to be constructive and get this review like a constructive one. Make hidden publicity is not so smart in the year what we are. I never blame you , i never disturb your offers (like you do) , i never accused you .... So wtf is your problem ? Be Jealous is not constructive believe me. Do not response to this comment because i know will not be nothing relevant and will came like "i am the big provider" because except of this you dont know to make nothing more. Enjoy.

    This thread is ignored start from now , i will not give him the pleasure ... i have been tagged so i have give my opinion.

  • Don't worry @cociu, I will always buy my Romanian servers from you. They smell nice and you take unwanted sisters as payment (who maybe don't smell so nice)

  • So @cociu got called out and proven inferior? I missed that thread. Can link me to the thread /post?

  • dedicatserver_rodedicatserver_ro Member, Host Rep

    After more than 25 years in the IT field ( yes i know , i´am old - 51 years old to be more precise ) I appreciate a lot :

    • Transparency and Honesty
  • tgltgl Member
    edited November 2019

    @cociu has another motto:

    Perfumes and Sisters

    Thanked by 1dosai
  • LizardFS... Holy sh** someone loves to be a blade runner. CEPH/ZFS is more common among storage providers. @dedicatserver_ro how long do you use LFS and what's your feedback on stability, disaster recovery (failed disk, electricity failures etc.)?

  • cociu said: Be Jealous is not constructive believe me

    Truth be told, that whole post makes you look pretty jealous. I get the coicu-speak and normally can translate to coherent but this one looks unhinged.

  • jsgjsg Member, Resident Benchmarker

    @EAgency said:
    The reviews I have seen from you are always done on instances that are specifically provided to you by the provider.

    Just like any sold system is specifically provided to customer. I mean, unless I do raids on providers with tanks and machine guns, I'll have to use what I'm provided with.
    Btw, I DO mention when a test system is different from the normal one. In fact, I did that in this very review! According to their web site the small dedi comes with 8 GB but my test dedi came with 16 GB - and I did expressly mention that.

    You also don't provide benchmarks for a reference system that is regularly provided and the provider doesn't know belongs to you.

    Wrong. I did quite a few benchmark/reviews on systems I purchased or got access to by a trustworthy source like a good friend.

    Without a reference system that has been provided like any other customers, you have absolutely no idea if the provider manipulates your testing environment in any form or shape f.e. by putting you on a node that is specifically selected or manipulated with artificial load and other shenanigans.

    Absolutely true - to a degree. Luckily though most providers are stupid enough to play tricks on me. If they did people would tell me that their system is far worse and I think being found out as a dirty trickster is something no provider likes.

    Also keep in mind that I'm not some kind of public institute with special rights (e.g. access) and providing official seals with grades. I'm just a guy who likes testing systems in a somewhat harder way than the usual nice number collector .benchmarks. I'm providing but one additional source of info - and it seems many people like my reviews and find them honest and helpful.

    If you give a provider the opportunity to provide you a system for testing there is always room for manipulation.

    The same is true with paying customers and with pretty much everything out there. Plus: If a provider plays dirty tricks and is found out, they will have a really hard time and loose customers.

    Also I think that as I'm not running a nice number collector but a real review it's less attractive (and easy) for providers to trick me. Also keep in mind that my benchmark is focused on "hunting" tricks (like e.g. aggressive caching).

    So you are basically providing a tiny snapshot of the performance of a VPS at the time it was provided to you by the provider on that specific node it's provisioned on at the specific time of your testing.

    Yes and no. Yes, my tests run over a short period of time, typ. 2 or 3 days, but No because they run at random times and at different times (e.g. day and night)

    Lastly talking about authenticity, how can I make sure you actually did test those products as someone who is interested in maybe renting from this specific provider you reviewed?

    I can just copy & paste stuff like that and change the values ...

    You can't. I fail to see however, why I would pull off a big fake benchmarking series.
    In my eyes it's simple: In a community everybody (well, ideally) tries to contribute something. Those benchmark and reviews are my contribution. If you don't trust me, simply do your own benchmarks.

    a) verify that you didn't get paid by the provider for this, especially since you did not spend any money yourself trying to get a "neutral" testing environment.

    Most of my reviews are not 100% positive or negative, because most products are not 100% positive or negative. I don't think that providers would pay for "so, so" reviews that also show weak spots.
    And btw, again, I'm providing a community service here. For free. Are you seriously expecting me to buy dozens of VPSs and dedis to be able to offer reviews that you would find more acceptable?

    b)verify the authenticity of your data because you don't provide me with proof you actually had access to those products and did test on them. Something like a YT Video or anything
    where you show that the system and IP belongs to the provider in charge and the actual testing procedures would give your data a lot more credibility

    Sure. And then a (payed for, of course) notary along with min. two (payed for) witnesses, verified cameras, background checked operators ....

    Like I said I respect all the effort you put into those reviews ...

    No, you don't and you made that very clear. But that's no problem. I can't make everyone happy (nor can anyone else). What p_sses me off though is that you stop short of calling me a liar and trickster - without any proof whatsoever, based only on wild speculations and "criticism". How funny to see Mr. "I only trust verifiable statements" trying to smear someone without any evidence.

  • jsgjsg Member, Resident Benchmarker

    @dedicatserver_ro

    All those numbers and graphics are nice and dandy but at the end of the day for a customer the results he gets count.

    What I'm way more interested in is the fact that you contacted me and indicated that you want to analyze eventual weak spots. That is what I'm more interested in and that is a positive attitude that deserves to mentioned here.

  • tgltgl Member
    edited November 2019

    well, to be fair this sounded more like a comparison than a review, I finally read this thread and I have concluded that: @dedicatserver_ro is more reliable and knowledgeable (we will see how @cociu will be when he will have 25 years of experience) although in business honesty and transparency aren't always going to help you, while @cociu is much cheaper, so it depends on what you want from a provider, on what you are using the service for

    if the european projects of @dedicatserver_ro will go through, he will probably offer better stuff in the near future, we will see if @cociu will start his own projects as he said in another thread

  • jsgjsg Member, Resident Benchmarker

    @tgl

    I of course know about the HS - DS story but no, this review is not a comparison. It's a review of two DS products. But I'll confess that the situation was my "trigger" to want to do a benchmark/review because frankly, DS's visibility is otherwise rather low here (rarely promos, etc.) so I like certainly many others know about DS mainly from their appearances in HS's threads.

    And in fact this review can not be a comparison because for that I would need to have tested a at least vaguely similar product but I actually tested only the current HS storage VPS which is not comparable but a different product.

    As for "HS is much cheaper", well, this its LET. Of bloody course price is important here.

  • tgltgl Member
    edited November 2019

    @jsg i understand what you mean, but it seemed that way because of the stuff you mentioned above, many people expected to see a comparison between the two because of the conflict, and they had just the cociu specs.

    one question remains though: how old is @cociu?

  • Around the age of 40.

    Thanked by 1tgl
  • jsg said: Just like any sold system is specifically provided to customer.

    With the difference that they know you are going to review their services specifically because you offer to. They know that whatever you write about them is having a positive or negative impact on their sales and even provide you with specs the customer doesn't get like you just admitted yourself.

    jsg said: According to their web site the small dedi comes with 8 GB but my test dedi came with 16 GB - and I did expressly mention that.

    That you mentioned that doesn't change that you made my point here exactly. You were provided with a system that is not what the paying customer gets. That does not make it an authentic user experience any longer and if the provider is willing to give you a better specced system it's not too far fetched he could also put you on a better node specifically because he knows you are reviewing his service and a very negative review has a huge impact on his sales potentially.

    jsg said: Also keep in mind that I'm not some kind of public institute with special rights (e.g. access) and providing official seals with grades. I'm just a guy who likes testing systems in a somewhat harder way than the usual nice number collector .benchmarks. I'm providing but one additional source of info - and it seems many people like my reviews and find them honest and helpful.

    How can you say you are not providing "seals and grades" but still try to pose an an authority and give clear recommendations and influence peoples buying decisions? That doesn't make sense at all.

    How is

    jsg said: Summary: Not recommended. Not as in "that's sh_tty", it's not, but as in "you'll get much better elsewhere for the money". As for the dedi, the system isn't bad, not at all, but it's simply way too expensive.

    not posing as an authority with a clear seal of approval? Seriously maybe I am just not seeing it because of my cancer meds, if I have a logical error somewhere let me know and I'll gladly apologize but do you give other users a clear recommendation and your (non)-approval or not?

    jsg said: You can't. I fail to see however, why I would pull off a big fake benchmarking series.In my eyes it's simple: In a community everybody (well, ideally) tries to contribute something. Those benchmark and reviews are my contribution. If you don't trust me, simply do your own benchmarks.

    Monetary gain. Free Servers. E-Peen. I've seen someone get stabbed over $20.
    Like I said I give you enough credibility to not do that but can I be sure? No. And that's also something you need to consider these days. When was the last time you shopped on Amazon and didn't see bought reviews even incl. videos and shit?

    You also never answered my question. What exactly is the value that you think you contribute by doing this in the way you do it? I mean you must know that because these are shared environments todays Benchmark doesn't need to be representative for next weeks Benchmark or the month after. You can be lucky and get a node with nice neighbors or on one that is completely oversold and new customers abuse the ressources given to them or one you get specifically put on by the provider because he knows you are going to judge his service publicly. Still at the end you try to give clear advise and a "seal of approval".

    jsg said: And btw, again, I'm providing a community service here. For free. Are you seriously expecting me to buy dozens of VPSs and dedis to be able to offer reviews that you would find more acceptable?

    I don't expect you to do anything, it's really nice you do this, I simply suggested if you want your reviews to have any weight at all and be useful it would be nice to have a reference system to compare with. You tell me you have friends who seem to buy from every provider you review and have those reference server(s), I don't understand why you are not making use of those extra ressources to put them in the review for more credibility but at the end of the day it's your reviewing process and your decision, but I also don't see why you get so butthurt over constructive criticism and the simple question how does your Benchmarks benefit me as a User on this website in making an adequate decision about buying something from a provider if you actually admit that Providers give you special test environments, you are not willing to verify your data in any way and then make statements like "Not recommended ( to buy )" to conclude your review in this case.

    jsg said: No, you don't and you made that very clear. But that's no problem. I can't make everyone happy (nor can anyone else). What p_sses me off though is that you stop short of calling me a liar and trickster - without any proof whatsoever, based only on wild speculations and "criticism". How funny to see Mr. "I only trust verifiable statements" trying to smear someone without any evidence.

    I don't see where any of my statements were disrespectful. If you can't take justified criticism of your process ( as I am not the only user who seem to have the same issues with it ) that is your problem, the only reason I took some time to tell you how you could actually make your reviews useful to someone wanting to make a buying decision with a certain provider is because I do respect the effort you put into this, but if you want to make claims about the quality of a service or provider and their services you should make sure people can actually trust the data you present and that your benchmarks are done on a configuration that is actually provided to normal customers under normal circumstances so I can get a better impression what I am actually buying into if I want to go with this provider.

    Also what are you 12? You tell me I "smear someone without any evidence" when I gave you totally reasonable arguments why it's hard to get any value out of your reviews and I am confused what exactly is your intention with them if they obviously can not be used to make an educated buying decision. Isn't that what a review is for? Finding out if a service is worth your money or not and what other real customers have as experience with a service provider?

    You completely refuse to give any evidence that your data is valid and even make fun of it, when it would be really easy to say hey this guy actually has a point and it doesn't hurt if I just record my testing sessions and upload them somewhere so people can be sure the data I provide is valid and I actually did test this provider. It's something that doesn't hurt you but gives your word even more credibility and doesn't cost you a notary, two witnesses and a camera. Brother you are doing an IT Service review, you are not doing a porn movie.

    Start free screen recording, end free screen recording when done, upload to free service. Done. 100% more credible JSG reviews.

    Like I said, don't be butthurt. You can take it as constructive criticism or tell me to fuck off. I just tried to be nice and help you make your very nice reviews actually useful for the people on this website.

    That's all from me. If you want to further discuss this you can always shoot me a PM. Otherwise have a great day and keep reviewing, I'll for sure have a look at all those nice numbers you post even if I don't gain anything from it in the end, I am German, i like looking at numbers, we're weird like that.

    Thanked by 1gogo
  • jsgjsg Member, Resident Benchmarker

    @EAgency

    I'll keep it short.

    • Your basic line of "can you prove?" is a classical way to attack. It is meant to put the other side at the loosing end because - particularly as I'm acting based on good will and without funding - I obviously can not prove and meet arbitrary and extreme criteria.
    • I've laid out my motivations and the context of operation clearly and transparently.
    • Your demands are ridiculous and mindless. Example: videos are not reliable proof anymore nowadays. Too much can be faked.
      Doing as nonsensically requested by you would only considerably increase my workload without really providing acceptable proof.

    Short version: I don't care AT ALL whether you trust and like my reviews. Just go and find someone else doing reviews the way you like them.

    I've had a lot of positive feedback and most of the readers trust that I try really hard to be fair and objective and that I know what I'm doing. I've earned the trust I enjoy here.

    Btw, you are not speaking for many. In fact you are the only one so far demanding for example to provide evidence on youtube.

    You've written 42 posts so far and quite a few of them are clearly anti @cociu, some of them even accuse cociu of selling fake watches. That you now appear in a thread reviewing a hoster who regularly sh_ts in cociu's threads and try to rip that not so great for that provider review apart seems to suggest your true motives.

    For the sake of fairness: I have no reason to think that dedicat_server has sent you. Their reaction so far was neutral to constructive.

    As for yourself I suggest you first contribute a bit more to our community and earn some standing before trying another dirty hit.

    Thanked by 2seriesn maverickp
  • C) cocks

  • @jsg said:

    @TimboJones said:
    Can you explain the random writes and why those numbers make a "solid SSD"? Maybe it's a formatting problem, but looks like 10.85MB/s. It would be better if you included headers with units, that's just basic benchmark and review stuff.

    • I already did explain that multiple times (and yes, the disk values are in MB/s)
    • I do understand that my way of doing it feels strange to some and I'm planning to develop a new version (sometimes, when I find time) with something that matches the usual outputs more.
    • but: My point was not to create yet another bla_bench providing nice but meaningless numbers.
      My point was to really check and benchmark systems, sometimes deliberately "breaking through" some common tricks like caches. I'm interested what a potential customer really gets and not in some impressive but largely meaningless numbers.

    • Yes, I confess it, my presentation is ugly and lacking, but hey, I'm a developer (not web!) and if it's not easy to format something nicely then I don't do it, simple as that.
      On the other hand I put a lot more effort in proper benchmarking and processing the result sets, as well as in interpreting them. And I'm investing typically quite some hours unlike someone who simply runs bla-test (twice if you are lucky) and is done. So, be a bit generous with me.

    Why no comment on the absolute shitty 10.85MB/s? (and spare the novel.) That was my point.

    I would argue you're not correctly interpreting the results. Nor is there any useful comparison with any other server. This is why every other type of review compares results against other products running same test, so it's meaningful and comparable. You're just spewing numbers for an OS that practically no one else will run. You're mistaken if you don't think you're not posting meaningless numbers.

    Holy shit, it takes you several hours to run script once? I didn't see where you mentioned averaging x number of tests or what time of day in the OP (and trying to say you test over days and nights over days makes me cringe, you'd be fired as a QA tester and laughed at for not providing basic, required test details with the results, instead you write linearly like you're doing this once). Spend your hours making a spreadsheet you can paste your numbers and get comparison graphs if you want it to be useful.

    Frankly, I'm just surprised you're so pleased with these mediocre "reviews". If you had made a site and called that a review, you'd get torn to shreds.

    Thanked by 1gogo
  • I almost feel bad for @jsg ...

  • uptimeuptime Member
    edited November 2019

    @EAgency said: I've seen someone get stabbed over $20.

    @jsg said: some of them even accuse cociu of selling fake watches

    Classic LET.

    Epic, really ...

    @deank has read the first 5 lines and wept - is now hiding all the kittens :smiley:

    Thanked by 2angstrom ITLabs
Sign In or Register to comment.