Howdy, Stranger!

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


One year of URPad and one important lesson learned
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.

One year of URPad and one important lesson learned

nerouxneroux Member
edited November 2013 in Reviews

A year ago I ordered three servers, one each in Buffalo, in Dallas, and in Los Angeles. The series of problems all started soon afterwards with LA becoming network-wise more and more unstable and with partial denial from their side and no solution in sight I eventually asked for a move from LA to Chicago (and was promptly moved to Dallas ;) ).

Somewhat later Buffalo suddenly started to behave funny and after dozens of ticket responses during the period of a week it got fixed and they established their eventual standard excuse of abusive customers. More on that later.

Next in line were incorrect system times, ranging from a few seconds to up to 20 minutes difference. It turned out they simply did not have any time synchronisation in place but obviously synced every now and then manually via rdate. I take the liberty and post some direct quotes

We have to use the rdate(get the time via the network) command to tweak the time.But we unable to know how it has happened.

By using rdate command we have had tweak the time settings.Actually whenever the node gets rebooted while the time also be changed.So in future will planned implementing some script that added the necessary modules once after the node gets rebooted.

If needed we will use that command also.But ntpd consumes some more resources.

Ok,We will implementing the script ASAP to prevent this issue in future.

We have forget to install ntpd thus,it has tweaked after rebooted the server.As of now i have installed on our server to prevent the issue in future.

After about 2.5 months they finally got the date issue sorted out, at least on the machines I was hosted on. However this was far from being done with problems as only then the previously mentioned abuser issues started.
In regular intervals (roughly at least once or twice a month) a machine became either entirely unreachable or showed an extremely poor performance. The usual procedure was first denial, then acknowledgement and then at some point later a brief "it works, please check" response without any further details. Quite often the issue was either not fixed at all or resurfaced a couple of hours late again.

At some point the performance in Buffalo and the level of support got so bad that I eventually requested a transfer to Dallas (where my existing instance worked for URPad's standards relatively trouble-free). Unfortunately this was not supposed to last long as the new instance showed right away an even worse performance. Because of this I eventually asked for yet another transfer to their (always highly advertised and praised) location in Houston, where the situation eventually improved somewhat (still sort of regularly unstable, but at least not constantly anymore).

All of this usually explained with abusive customers on the same host, regular denial of service attacks or simple power outages (no UPS?), garnished with traceroute or ping excerpts when you'd least expect them.
What should be also mentioned is that at least on one occasion they logged into the server, respectively accessed the file system without my consent.

Last but not least a recent response which could explain a lot of their issues

Regarding the server monitoring, we aren't monitoring the i/o,cpu resources usage

The forced move now (they close down all of their datacenters and move to three locations only) was the final straw, after already being forced to move three times to compensate for their hardware and competence issues, I decided to cancel the two accounts in question. The entire procedure took two full days, several enquiries from my side asking for confirmation of the terms, several unrelated responses from their side and eventually the cancellation of all three services (even the paid one) without any notice whatsoever (what I actually specifically asked for from the beginning). Fortunately I took a backup beforehand, anticipating this.

With one exception (when Mr. Armado was quite snippy) their support was usually courteous but unfortunately this compensated neither for the support's general slowness nor its, I am afraid to put it so bluntly, simple and utter incompetence (Curtis Manning was a rare exception).

If there is one thing I learned from being a URPad customer and which I could share as experience, then - unless you know a service provider really really well - do not ever ever ever pay in advance for a full year.

Thanked by 1FrankZ

Comments

  • rm_rm_ IPv6 Advocate, Veteran
    edited November 2013

    do not ever ever ever pay in advance for a full year.

    Amen to that :)

    unless you know a service provider really really well

    I'd say "and even then". Situations and requirements change, new good deals pop up all the time, and you invariably will find yourself waiting on some VPS you no longer need to expire in a few months (and lament how you wasted those money).

    And also I'd add never ever use OpenVZ (the time problems you described are so widespread it's not even funny). Also I assume you were lucky to not need TUN/TAP or any additional iptables modules, judging from the host's competence it could have taken an equal amount of time to get THAT loaded/enabled/persist across reboots.

  • with LA becoming network-wise more and more unstable

    This is not necessarily a problem with URPad. I regularly see major network issues in and around LA. I've seen these network problems occur with quality providers like INIZ, CVPS, DotVPS, and TeraFire. Los Angeles network issues are to be expected.

  • Did you just place cvps on the same level as iniz?

  • @joelgm said:
    Did you just place cvps on the same level as iniz?

    They both use nLayer in LA, both offer SSD in LA, and they both have around 4000 iops in LA. So, yeah, I guess I did. ;-)

  • Urpad's L.A node rocks, faster ping to Asia and S. America, very stable.

    Thanked by 1ChrisMiller
  • nerouxneroux Member
    edited November 2013

    @jimpop said:
    This is not necessarily a problem with URPad. I regularly see major network issues in and around LA. I've seen these network problems occur with quality providers like INIZ, CVPS, DotVPS, and TeraFire. Los Angeles network issues are to be expected.

    As you can see from the thread LA was not the only issue with URPad nor the worst.

    I cannot comment on other providers but I'd be surprised if it really only depended on the location itself. But even if this were the case, it would still fall into URPad's responsibility to choose (or not to choose) such a location.

    @MorningIris said:
    Urpad's L.A node rocks, faster ping to Asia and S. America, very stable.

    In this case unfortunately not. It rather looked like if all of California was connected via one single 150 bps acoustic coupler.

  • do not ever ever ever pay in advance for a full year.

    Preferable, for me. LEB providers give low end and hopefully as little support required as possible, but the time overhead applies to me too. I don't want to mess around with monthly invoices, and having subscriptions is easier but still a hassle to manage.

  • Wow.Urpad is not bad.But they some nodes is not well, some nodes is great.ticket speed is slow, I am very lucky,my L.A node is not bad.

    dd if=/dev/zero of=test bs=64k count=512 conv=fdatasync
    512+0 records in
    512+0 records out
    33554432 bytes (34 MB) copied, 0.239407 s, 140 MB/s

    dd if=/dev/zero of=test bs=64k count=512 oflag=dsync
    512+0 records in
    512+0 records out
    33554432 bytes (34 MB) copied, 0.796441 s, 42.1 MB/s

    wget http://cachefly.cachefly.net/100mb.test
    --2013-11-09 00:38:30-- http://cachefly.cachefly.net/100mb.test
    Resolving cachefly.cachefly.net... 205.234.175.175
    Connecting to cachefly.cachefly.net|205.234.175.175|:80... connected.
    HTTP request sent, awaiting response... 200 OK
    Length: 104857600 (100M) [application/octet-stream]
    Saving to: “100mb.test”

    100%[======================================>] 104,857,600 39.1M/s in 2.6s

    2013-11-09 00:38:33 (39.1 MB/s) - “100mb.test” saved [104857600/104857600]

    Thanked by 1ChrisMiller
  • @neroux - Thank you for taking the time to write this review.

  • @neroux said:

    A year ago I ordered three servers, one each in Buffalo, in Dallas, and in Los Angeles.

    ... (snipped tale of woe)

    If there is one thing I learned from being a URPad customer and which I could share as experience, then - unless you know a service provider really really well - do not ever ever ever pay in advance for a full year.

    The lesson I would've learned from this story is: don't put all your eggs in the one basket.

  • @neroux

    said: If there is one thing I learned from being a URPad customer and which I could share as experience, then - unless you know a service provider really really well - do not ever ever ever pay in advance for a full year

    I can say to you: I learnt a lot more with my WORST PROVIDER experience with URPAD (in fact it was K-disk and after they migrated to URPad).

    Why can we not see good reviews (except the paid one in WHT) about URPAD?

    And I can not understand WHY people try them, yet.
    Maybe the price!
    Maybe the template used in its INDEX folder.

    Thank you for your review.

  • concerto49concerto49 Member
    edited November 2013

    @jimpop said:
    This is not necessarily a problem with URPad. I regularly see major network issues in and around LA. I've seen these network problems occur with quality providers like INIZ, CVPS, DotVPS, and TeraFire. Los Angeles network issues are to be expected.

    It's more like ALL 1 single provider - Colocrossing and/or Quadranet. It's not an LA issue. You're talking about 1 data center. The issue is NOT LA.

  • budi1413budi1413 Member
    edited November 2013

    concerto49 said: Quadranet

    I agree Quadranet LA network is not good anymore. I'm happy with WebNX now.

  • @concerto49 said:
    It's more like ALL 1 single provider - Colocrossing and/or Quadranet. It's not an LA issue. You're talking about 1 data center. The issue is NOT LA.

    The issue in LA is LA tranist providers, not datacenters. Anyone running traces can see this.

  • The issue with LA (or west coast in general) is that it's close to China.

  • c0yc0y Member
    edited November 2013

    Zen said: I won't name names, but I know of a big UK provider who's technical team and himself refuse to believe that a BGP null route doesn't continue to cost them in bandwidth after applied

    Do you mean Rapidswitch?

  • @Zen said:
    WHISTLES

    (:

  • @neroux Thank you for the feedback, This really does help us. Since Root Level Technology has taken over, it has been a goal to get all systems running to their full potential. We have cleaned several servers up, and now fight abusers on a daily basis to ensure that regular customers do get the product they were promised. Since the time that someone has told you that we do not monitor CPU and I/O, that has changed. We monitor all network traffic, ping times, CPU, I/O, Disk space, server and network hardware temperatures, and also are implementing several scripts to ensure that time and other issues are resolved.

    In regards to the issues with Mr. Armado, We have addressed them and are working on ensuring the utmost care is put into every reply. We have also brought on several new support staff to ensure that response times are met.

    If you do have skype, please add me at the username below. I would like to hear more of your feedback on ways that we can improve.

    @MorningIris Thank you for the positive comments.

    @zxsdw Right now we are in the process of cleaning all the servers up, and adding new servers to balance out those that are a bit slower.

    @sleddog We do offer most of our promotions in a yearly term due to the transactions fees, and etc. But we are more than willing to work with anyone on a monthly basis. Its my firm belief that a product should speak for its self. A lot of providers out there make you sign long term based contracts, If they were confident in their service and support they would not make you do that.

    As always, I would like to invite anyone to add me on skype if you have questions or concerns as I would like to hear them. Skype name is RandalBurnsRLT

    Thanks!

    Thanked by 1Lee
  • @rm_ does your aversion to paying yearly apply to dedis too?

  • rm_rm_ IPv6 Advocate, Veteran
    edited November 2013

    craigb said: does your aversion to paying yearly apply to dedis too?

    Sure, with dedis doubly so, since they tend to be more expensive, the total amount paid ends up being huge, and in case of any problems much more is at stake than with some VPS for $15/year. Providers like digicube.fr really do tempt you to pay yearly for a dedi (by giving a 20% discount), but those savings may end up being nullified by a case when some five months down the road you find a better deal, or the service quality goes to shit, or you simply do not need that dedi anymore, etc.

    P.S.: I am currently waiting for my yearly dedi from OVH to expire, already do not need it since 2-3 months ago, and it's still got about one month left...

    Thanked by 1craigb
  • concerto49concerto49 Member
    edited November 2013

    @jimpop said:
    The issue in LA is LA tranist providers, not datacenters. Anyone running traces can see this.

    No it isn't. Transit carriers always have disputes somewhere and always have congestion somewhere. It's up to the data center and service providers to have many different carriers in the mix with intelligent routing to work around these problems.

    I have heard Quadranet are going to introduce NTT soon, so hopefully this will fix it. I'm sure Colocrossing is pushing them hard :P

  • Urpad is one of the only low-end providers here that won't be disappearing any time soon. For a VPS at $7/Month or under $7/Month, expect these issues. $7/Month is not much profit what so ever, I am sick of people complaining about services at or under $7/Month, it's not like each of these VPS servers cost $20/Month for 512MB Ram. I've used Urpad multiple times within my years of experiencing OpenVZ VPS servers and they never let me down, once. Chris is a great guy, and so is the company that bought Urpad.

    Thanked by 1Lee
  • @concerto49 said:
    No it isn't.

    @concerto49 said:
    Transit carriers always have disputes somewhere and always have congestion somewhere.

    So... which one is it?

  • @jimpop said:
    So... which one is it?

    Transit carriers always have problems, but the problems the consumer is seeing isn't their fault. They should be oblivious to what goes on in the background. I'd like to ask who you are though?

  • jimpopjimpop Member
    edited November 2013

    @concerto49 said:
    Transit carriers always have problems

    Which is exactly the point I made, that you countered. You seem to think that low-end providers are going to have multiple paths of connectivity which automatically adjust to loss and peak usage. lol.

    I'd like to ask who you are though?

    I'm a guy who knows networking. ;-)

    Edit: and I've also paid for at least a dozen VPSes, across multiple datacenters, in LAX within the past year.

  • @jimpop said:
    Edit: and I've also paid for at least a dozen VPSes, across multiple datacenters, in LAX within the past year.

    Great stuff :)> @jimpop said:

    Edit: Which is exactly the point I made, that you countered. You seem to think that low-end providers are going to have multiple paths of connectivity which automatically adjust to loss and peak usage. lol.

    For the prices that the dedicated servers go for, why not? A lot of things can be done at the price point - well not the $1/year 4GB KVM offers of course. Some price points have killed the competition. Besides, you don't need a lot to adjust.

    Considering we sell these things too, I'm telling you it can be done :)

  • jimpopjimpop Member
    edited November 2013

    @concerto49 said:
    Considering we sell these things too, I'm telling you it can be done :)

    You automatically (unmanned, 24x7x365) auto-negotiate optimized pathing for ingress and egress traffic on the networks you administer for your LET/LEB VPS clients?

    Edit: added LET/LEB

  • @jimpop said:
    You automatically (unmanned, 24x7x365) auto-negotiate optimized pathing for ingress and egress traffic on the networks you administer for your LET/LEB VPS clients?

    I said it can be done. That's what products like Noction and Internap FCP do. A lot of providers here have it, including those that sell on LET/LEB. Not everything can be automated. It's not god mode, but definitely helps.

  • That's what I thought. ;-)

Sign In or Register to comment.