Howdy, Stranger!

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


Why do I stay at hostsolutions.ro?
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.

Why do I stay at hostsolutions.ro?

Hello everyone,

I apologize for my bad English, as I only speak French.

Why do I stay at hostsolutions.ro?

Why do I like hostsolutions.ro?

They are not perfect, sometimes there are problems, a few years ago I had a VPS that totally

crashed, and more recently there was a big problem in their datacenters in Romania.

However, for me they are 200% reliable!

They know when there are problems and they always do their best to solve them.

They have an incredible professional conscience, they are accessible, they have a human dimension.

The founder of the company is always present on this forum to answer all our questions, to find solutions.

Yes, they are not perfect, but which company is 100% perfect?

I have never regretted being with them, even if sometimes I was frustrated by breakdowns or cuts.

They have excellent service and great value for money.

I sincerely recommend hostsolutions.ro

Thanked by 2_Nic pulseone

Comments

  • tolovalltolovall Member
    edited December 2020

    @cociu, you get some fans. Don't let them down.

    Thanked by 2cociu maverickp
  • @comeback said:
    Yes, they are not perfect, but which company is 100% perfect?

    buyvm and knowhost

    Thanked by 1chip
  • yoursunnyyoursunny Member, IPv6 Advocate

    @comeback said:
    Yes, they are not perfect, but which company is 100% perfect?

    summer host is perfect. YABS comes pre-installed.

    @SmartHost is also perfect. Zero downtime (that I noticed) in two years.
    However, I'm canceling next year because I found something cheaper and have lower latency to my readers.

    Thanked by 1ariq01
  • Fully agree with you

    @comeback said: They know when there are problems and they always do their best to solve them.

    They have an incredible professional conscience, they are accessible, they have a human dimension.

    The founder of the company is always present on this forum to answer all our questions, to find solutions.

    Thanked by 1cociu
  • @comeback said: comeback

    lol man , really thanks i have need such thread after been fucked the entirely year , this year was the most difficult one for me in this business (a lot of problems, grow like a idiot and uncontrollable, new locations , etc etc. police raid , downtimes) but hey we are here. Yes we are not perfect , never will be , but what i know all this year we have inverted in network (for this was many downtime too) and i have a target quality for 2021 wich i will do my life to get it. Any way thanks again and i wish you and for all Happy 2021.

  • raynorraynor Member
    edited December 2020

    Have 3 storage and 1 SSD (yes i was bf-crazy) VPSs now, so can add me to fans of dirt-chaep storage with fun and sistars! well and english-like wizardry ovc. btw 2 VPS is over 1 year in backup job and I must say that I expected much worse quality than I have now - some connection issues solved by tikats and full data loss due to centos8 template (Debian/FreeBSD only since this), but it's all :)

    Thanked by 1cociu
  • @raynor said: Have 3 storage and 1 SSD (yes i was bf-crazy) VPSs now, so please add me to fans of dirt-chaep storage with fun and sistars! well and english-like wizardry ovc.

    thanks!

  • @comeback said:

    Yes, they are not perfect, but which company is 100% perfect?

    OVH

  • @comeback said:

    Why do I stay at hostsolutions.ro?

    Why do I like hostsolutions.ro?

    I sincerely recommend hostsolutions.ro

    To thank @cociu , just post more sisters.jpg :#

    Thanked by 1cociu
  • This is the weirdest thread I've seen for a long time.

  • jsgjsg Member, Resident Benchmarker

    I wouldn't go as far as OP, but I see that reliability and availability also have a counter-part on our side. We simply very often do not need high reliability and availability, even many small businesses don't need 100% (anyway largely a wet dream) or 99.995% but can live well with 99.9% or even less.
    Example: a backup server. it's down for 2 minutes or hours or even days? So what, (a) if you rely on one backup server only and from one provider only and in one location only then the major problem is on your side anyway, and (b) a good backup solution works similar to SMTP, it tries again and again ... and finally succeeds.

    Of course I wouldn't advise anyone, business or private, needing reliability and availability north of 99% to get a HostSolutions system - but frankly we, the other 80% of customers in at least 80% of use cases actually can live well with 99% availability - especially when that also means considerably lower cost, often up to 50% (and even more) cheaper.

    Now that, so it seems, @cociu has left the "a ticket? oh well, maybe you'll have an answer in 5 hours, maybe in 5 days" nightmare phase behind themselves, I find it reasonable to wish for cociu aiming at achieving a reasonable next step which could be to achieve 99.5% availability and ticket response times of less than say 3 hours (8 am to 8 pm) and say less than 6 hours on weekends.

    Oh and please do not target beyond 99.9% because I want you to stay cheap. It's all a question of balance and finding a sweet spot on the quality - cost axis.

    Thanked by 3raynor cociu _MS_
  • yoursunnyyoursunny Member, IPv6 Advocate

    @jsg said:
    I wouldn't go as far as OP, but I see that reliability and availability also have a counter-part on our side. We simply very often do not need high reliability and availability, even many small businesses don't need 100% (anyway largely a wet dream) or 99.995% but can live well with 99.9% or even less.

    99.995% => 30s downtime per week, or 26m17s downtime per year
    99.9% => 10m4s downtime per week, or 8h45m56s downtime per year

    I can live with that.
    However, the police raid caused several days of downtime for some servers.

    Example: a backup server. it's down for 2 minutes or hours or even days? So what, (a) if you rely on one backup server only and from one provider only and in one location only then the major problem is on your side anyway, and (b) a good backup solution works similar to SMTP, it tries again and again ... and finally succeeds.

    98% => 3h21m36s downtime per week, or 7d7h18m59s downtime per year

    It's OK for backup storage, but too much for a website.
    If my website is down for more than 12 hours, I'll start deploying the website on another server. It only takes an rsync from the laptop.

    Thanked by 1lentro
  • Anybody else got HS VPSs for websites? :D

  • jsgjsg Member, Resident Benchmarker
    edited December 2020

    @yoursunny said:
    However, the police raid caused several days of downtime for some servers.

    Yes but I consider that force majeur and it could happen to any provider.

    Example: a backup server. ...

    98% => 3h21m36s downtime per week, or 7d7h18m59s downtime per year

    It's OK for backup storage, but too much for a website.

    Depends, mainly on (a) type of website (e.g. private vs. say a web-shop), and (b) on the downtime spread. Multiple down-times of some minutes is OK, multiple hours (every month) highly likely is not.

    If my website is down for more than 12 hours, I'll start deploying the website on another server. It only takes an rsync from the laptop.

    ... or from a backup or deployment server. But keep DNS in mind which can throw a stick into the gears. And of course while you have something in place many (most, I guess) do not.

    But anyway I consider 99% avail. (daily: 14m 24s ... monthly: 7h 18m 17s) as a reasonable minimum for most private use cases and in fact try to not get below 99.5%.

    Does anyone have current avail. data of HostSolutions (ignoring of course the police/SRI raid)?

  • edited December 2020

    @jsg said:
    Does anyone have current avail. data of HostSolutions (ignoring of course the police/SRI raid)?

    Mine have 96.94% uptime and 96.89% uptime within 60 days range with downtime around 44 hours

    EDIT: Ignoring police raid with 30 hours of down, the uptime should be around 99% to 99.1%

    Thanked by 2jsg cociu
  • Ovh

  • cociucociu Member
    edited December 2020

    @chocolateshirt said: Mine have 96.94% uptime and 96.89% uptime within 60 days range with downtime around 44 hours

    next week we will have another downtime in oradea (multiple ones in the same day but for short period) after this you can start count and we will talk in 1 year :)

    Thanked by 1chocolateshirt
  • jsgjsg Member, Resident Benchmarker

    @cociu said:
    next week we will have another downtime in oradea (multiple ones in the same day but for short period) ...

    Reason, please

  • cociucociu Member
    edited December 2020

    @jsg said: Reason, please

    missing some stuff to have all fully redundant , even buckle of fibers , routers , swich . So missing like 30 min intense job in order to start 2021 like i always wanted.

    sorry i am not a fiber man or networking one so more details via tiketing with att Adrian or Andrei and i am sure will be glad to explain.

    Thanked by 2jsg _Nic
  • @cociu said:

    @jsg said: Reason, please

    missing some stuff to have all fully redundant , even buckle of fibers , routers , swich . So missing like 30 min intense job in order to start 2021 like i always wanted.

    sorry i am not a fiber man or networking one so more details via tiketing with att Adrian or Andrei and i am sure will be glad to explain.

    any chance to get some pictures of the Oradea location ? :)

  • @Edding said: any chance to get some pictures of the Oradea location ?

    this year i have invited and received 2 persons from LET so if want to share the experience (ony with a write presentation) i dont have any to hide.

    i am not adept of pics , for sure i am not a huge international company , we are in a warehouse and we have until now 38x48u racks full and some iddle waiting to be populated. All what was until 2 years ago was custom made but in the last 2 years (more precise when we have change the location to a bigger space) i have invert much in imagen tooo so we have change many things here. Still need improove much but hey this is my target now , thanks to all customers we have a good income in a difficil period so only need time and i hope this business will be like anyone want it.

    PS i invite any person from here to take a visit (2 days hotel +dinner will be invited) but no for the moment we are not for pictures sorry.

    Thanked by 2Edding _Nic
  • jsgjsg Member, Resident Benchmarker

    @cociu said:

    @jsg said: Reason, please

    missing some stuff to have all fully redundant , even buckle of fibers , routers , swich . So missing like 30 min intense job in order to start 2021 like i always wanted.

    sorry i am not a fiber man or networking one so more details via tiketing with att Adrian or Andrei and i am sure will be glad to explain.

    No worries, that's good enough - and it's a very positive reason too.

    Best wishes that everything works as planned.

  • yoursunnyyoursunny Member, IPv6 Advocate

    @raynor said:
    Anybody else got HS VPSs for websites? :D

    I don't have HostSolutions VPS. I hear the SolusVM is broken that there's no way to reinstall. This is the major reason that I would not purchase, since I need to reinstall several times to work out the deployment procedure.
    Apart from that, SSD plans are for websites and APIs and push-up video repositories.


    @jsg said:

    @yoursunny said:
    However, the police raid caused several days of downtime for some servers.

    Yes but I consider that force majeur and it could happen to any provider.

    True. It even happened to NexusBytes Germany two months ago.
    However, this doesn't change the fact that SLA has been violated.

    Example: a backup server. ...

    98% => 3h21m36s downtime per week, or 7d7h18m59s downtime per year

    It's OK for backup storage, but too much for a website.

    Depends, mainly on (a) type of website (e.g. private vs. say a web-shop), and (b) on the downtime spread. Multiple down-times of some minutes is OK, multiple hours (every month) highly likely is not.

    We are losing billions per hour or millions per minute!

    12 years ago, I was in charge of distributing study materials to classmates through my website.
    If the website is down for more than an hour, classmates start knocking on my door.

    My current readers do not personally know me, but I feel bad if the website becomes inaccessible for too long.

    If my website is down for more than 12 hours, I'll start deploying the website on another server. It only takes an rsync from the laptop.

    ... or from a backup or deployment server. But keep DNS in mind which can throw a stick into the gears. And of course while you have something in place many (most, I guess) do not.

    Don't everyone have a few idling servers?

    Speaking of DNS, I got an idea: I could always rsync two copies of the website in different locations. The secondary server periodically checks the primary. If the primary is down, the secondary modifies DNS via Cloudflare API.

    Background information: my main website is using Cloudflare DNS but not proxy. I have 40% readers in China and Cloudflare proxy is too slow for them.
    rsync works for me because the website has no database. The content is entirely determined by the files and external APIs.
    Caddy web server can automatically obtain TLS certificate when needed; or I can rsync the certificate too.

  • @cociu said:

    @Edding said: any chance to get some pictures of the Oradea location ?

    this year i have invited and received 2 persons from LET so if want to share the experience (ony with a write presentation) i dont have any to hide.

    i am not adept of pics , for sure i am not a huge international company , we are in a warehouse and we have until now 38x48u racks full and some iddle waiting to be populated. All what was until 2 years ago was custom made but in the last 2 years (more precise when we have change the location to a bigger space) i have invert much in imagen tooo so we have change many things here. Still need improove much but hey this is my target now , thanks to all customers we have a good income in a difficil period so only need time and i hope this business will be like anyone want it.

    PS i invite any person from here to take a visit (2 days hotel +dinner will be invited) but no for the moment we are not for pictures sorry.

    Register your account at Instagram. The username 'hostsolutionsro' is still available. Many other VPS/Dedi providers that have their own DC/space post pictures of servers and promos.

  • @MS said: Register your account at Instagram. The username 'hostsolutionsro' is still available. Many other VPS/Dedi providers that have their own DC/space post pictures of servers and promos.

    not interested in social media ... i am adicted to LET still ...

    Thanked by 2brueggus tux
  • jsgjsg Member, Resident Benchmarker

    @yoursunny said:

    Depends, mainly on (a) type of website (e.g. private vs. say a web-shop), and (b) on the downtime spread. Multiple down-times of some minutes is OK, multiple hours (every month) highly likely is not.

    We are losing billions per hour or millions per minute!

    I know, I know, I would have taken over OVH long ago if I hadn't lost many billion $$ due to a VPS that was down for 7 or 8 minutes.

    12 years ago, I was in charge of distributing study materials to classmates through my website.
    If the website is down for more than an hour, classmates start knocking on my door.

    Yep, that matches with my experience too. Occasionally 15 - 30 min offline is a nuisance but not a problem. Once downtime is over 30 min people get irritated and start to complain.

    If my website is down for more than 12 hours, I'll start deploying the website on another server. It only takes an rsync from the laptop.

    ... or from a backup or deployment server. But keep DNS in mind which can throw a stick into the gears. And of course while you have something in place many (most, I guess) do not.

    Don't everyone have a few idling servers?

    I didn't mean a spare VPS but rather a solution approach at all (like e.g. your rsync mechanism).

    Speaking of DNS, I got an idea: I could always rsync two copies of the website in different locations. The secondary server periodically checks the primary. If the primary is down, the secondary modifies DNS via Cloudflare API.

    The point isn't synching but rather DNS propagation delay plus the fact that many provider don't care sh_t about the rules (and your zone config).

    ... Cloudflare ...

    I do not use any CloudF§#&! services at all, just like I wouldn't use a free NSA service.

    Thanked by 1yoursunny
  • @comeback said:
    They are not perfect, sometimes there are problems, a few years ago I had a VPS that totally

    crashed, and more recently there was a big problem in their datacenters in Romania.

    However, for me they are 200% reliable!

    And nobody should be giving a fuck what your opinion is based on this.

  • Why do I ?
    Love the way you only always leave me
    Love the touch but never love the feeling.

  • lentrolentro Member, Host Rep

    @yoursunny said: I don't have HostSolutions VPS. I hear the SolusVM is broken that there's no way to reinstall. This is the major reason that I would not purchase, since I need to reinstall several times to work out the deployment procedure.

    @cociu uses Proxmox + this ModulesGarden plugin. Works just fine, although reinstalls take quite a bit of time.

    @cociu said: 38x48u racks full

    Btw - @cociu did some math, you must be making bank! Good luck!

  • yoursunnyyoursunny Member, IPv6 Advocate

    @cociu said:
    PS i invite any person from here to take a visit (2 days hotel +dinner will be invited) but no for the moment we are not for pictures sorry.

    dat centr tour, wow!


    @lentro said:

    @yoursunny said: I don't have HostSolutions VPS. I hear the SolusVM is broken that there's no way to reinstall. This is the major reason that I would not purchase, since I need to reinstall several times to work out the deployment procedure.

    @cociu uses Proxmox + this ModulesGarden plugin. Works just fine, although reinstalls take quite a bit of time.

    Good to know.
    When I finish setting up Black Friday boxes, I'll consider.


    @jsg said:

    Don't everyone have a few idling servers?

    I didn't mean a spare VPS but rather a solution approach at all (like e.g. your rsync mechanism).

    I write content on the laptop, push them to the server via rsync, and separately push the source to Bitbucket (it's public repository, but fewer people are looking at my Bitbucket than my GitHub account).
    I know the better way is to push to source control and let CI deploy to the server, but I feel it's too slow, so I haven't adopted this model.
    Either method could upload to multiple servers.

    I know some webmasters like to write content on the webpage into a database (ahem, WordPress and MediaWiki).
    This is not good because there's always a chance that the database crashes right after you submit the content but before automatic backup occurs.
    You can setup realtime replication but failover procedure is tricky to get right, unless you pay a major cloud provider to do that for you.

    Speaking of DNS, I got an idea: I could always rsync two copies of the website in different locations. The secondary server periodically checks the primary. If the primary is down, the secondary modifies DNS via Cloudflare API.

    The point isn't synching but rather DNS propagation delay plus the fact that many provider don't care sh_t about the rules (and your zone config).

    You can select a good provider for authoritative name server, that updates fast and responds with correct TTL.
    Then, major recursive DNS providers (Google, Cloudflare, OpenDNS, etc) do respect TTL. Ignore the crappy providers.

    Suppose I can accept a downtime of up to one hour, I would set TTL to half hour. If the primary web server has been down for half hour, I change A/AAAA records.

    ... Cloudflare ...

    I do not use any CloudF§#&! services at all, just like I wouldn't use a free NSA service.

    Cloudflare Registrar is the cheapest, so they control all my domains except .cn.
    Occasionally a domain gets 1-year sleepovers at the Porkbun or NameCheap guesthouse, when there's a transfer discount.

    However, I turn off the CDN/proxy for the main site.
    Other sites has the CDN/proxy turned on so that there are fewer TLS certificates to management, but I don't use these annoying CAPTCHAs.

    If NSA offers a free VPS, I would use it to store encrypted files.
    If NSA offers a free mobile phone with free data plan, I would use it to record push-ups and watch TV, but probably won't bring it with me at all times.

    Thanked by 1lentro
Sign In or Register to comment.