Results 1 to 24 of 24
  1. #1

    Wiredtree Vs Servint - Dedicated Server Configuration Options

    Hey experts! Your thoughts on which server option below is best will be appreciated.

    I have quotes from Servint and Wiredtree for fully managed NON-Virtuozzo dedicated servers with exactly SAME COST but with different configuration. The server will be used to host 200 small-medium domains with many (strictly-no-spam) mail accounts and a few small databases.

    Common things offered by both are:

    Quad Core Xeon® X3440 Processor
    4 x 2.4 GHz Hyperthreaded Cores
    8 MB L3 Cache, 1333 MHz FSB
    4 GB 1333MHz DDR3 ECC Memory

    Differences are:

    A) Disks and Raid controller :-

    SERVINT
    2 x 146 GB "10K" SAS HDD in Hardware RAID 1
    SAS i6R SAS internal RAID adapter PCIe - 256 MB cache

    WIREDTREE
    2x147GB "15K" RPM SAS 16MB Cache
    Adaptec 5405 PCIe - 1.2GHZ Dual Core - 256MB Cache
    Battery Backup Unit (BBU) for Raid 1

    B) Backup :-

    SERVINT
    EITHER - Backup Drive 500GB 7.2k RPM SATAII 16MB Cache
    "OR" - 50GB R1soft backup within the DC itself

    WIREDTREE

    - Backup Drive 500GB 7.2k RPM SATAII 16MB Cache
    "AND" - 50 GB remote FTP backup space

    Wiredtree also offers Servershield hardening for free.

    Both are reputed for support and I know both are good companies. My Queries are:

    1. Since cost is exactly same, which server above is the clear winner and why?

    2. Have never felt the need to restore a single file, just want the backup incase of a crash. Is just a backup drive not enough?

    3. In the Wiredtree backups, can both backup drive and Remote FTP backup be useful on a cPanel server? If not should I save $25 monthly and let go of remote backup since local allows quicker restore?

    4. Even with BBU, if battery fails there is a very very small data risk if write caching is enabled. Should I save $20 monthly on Wiredtree server by avoiding the Battery Backup Unit for RAID1 and disable write-caching?

    5. From my location, average ping time for Servint is 310ms and Wiredtree is 330ms. Will 20ms be a perceptible difference for websites and emails?

    Many thanks in advance for your advice.
    Last edited by aarkay777; 06-03-2011 at 11:09 AM.

  2. #2
    Join Date
    Nov 2009
    Location
    Cincinnati
    Posts
    1,585
    The Servint setup is a dell 6ir sas controller, no write cache. You need to find out if the 5405 card from wiredtree has write cache and if it has the BBU.

    Wiredtree has had a lot of network issues/outages the past 2 months, not sure whats causing that.

    I've seen countless threads about it. Normally I'd say wiredtree hands down comparing the 2.
    'Ripcord'ing is the only way!

  3. #3
    Join Date
    Oct 2003
    Location
    Chicago, IL
    Posts
    657
    Quote Originally Posted by Visbits View Post
    The Servint setup is a dell 6ir sas controller, no write cache. You need to find out if the 5405 card from wiredtree has write cache and if it has the BBU.

    Wiredtree has had a lot of network issues/outages the past 2 months, not sure whats causing that.

    I've seen countless threads about it. Normally I'd say wiredtree hands down comparing the 2.
    No, actually we haven't had any network issues or outages in the last 2 months. You must be confusing us with someone else. We've made a number of large network upgrades in the past two months, but haven't suffered any outages whatsoever.

    The Adaptec 5405 has 512MB cache and we offer the optional BBU on our site as well.
    Zac Cogswell / CEI
    Formerly known as WiredTree Zac

  4. #4
    Join Date
    Jun 2009
    Posts
    1,219
    From my personal experience with both providers, I would choose WiredTree. Also I would always prefer 15k SAS drives compared to 10k HDDs.

  5. #5
    Quote Originally Posted by Visbits View Post
    The Servint setup is a dell 6ir sas controller, no write cache. You need to find out if the 5405 card from wiredtree has write cache and if it has the BBU.

    Wiredtree has had a lot of network issues/outages the past 2 months, not sure whats causing that.

    I've seen countless threads about it. Normally I'd say wiredtree hands down comparing the 2.
    1) As mentioned in my original post, the Wiredtree server has write cache and BBU. Hence, I had asked this above:-

    "4. Even with BBU, if battery fails there is a very very small data risk if write caching is enabled. Should I save $20 monthly on Wiredtree server by avoiding the Battery Backup Unit for RAID1 and disable write-caching?"

    2) I researched before posting here but could not find any recent outage threads for Wiredtree or Servint. Do you have links or was it another company you were mistakenly referring to?
    Last edited by aarkay777; 06-04-2011 at 04:59 AM.

  6. #6
    Join Date
    Dec 2004
    Location
    New York, NY
    Posts
    10,710
    Write caching can improve performance quite a lot. It is worth the extra cost for the BBU.

  7. #7
    Quote Originally Posted by layer0 View Post
    Write caching can improve performance quite a lot. It is worth the extra cost for the BBU.
    Thanks boss. Happy to pay an extra $20 / mo for this.

    What about the small data risk in case the BBU battery or DC power backup goes down?

    Also for 200 medium domains with lots of email accounts will the write caching speed advantage be perceptible?

  8. #8
    I seem to be the only one having fun with my own thread.

    So does everyone recommend taking the small risk of write caching with a BBU in place?

    Also, average ping time for Servint is 310ms and Wiredtree is 330ms from my location. Will 20ms be a perceptible difference for websites and emails?

  9. #9
    Join Date
    Mar 2003
    Location
    Kansas City, Missouri
    Posts
    462
    Hello,

    Write cache is a must for systems utilizing database engines like MySQL. Cache memory is always faster than disk. BBU keeps that cache persistent if power is lost. When the system is rebooted, cache is replayed to disk and you have consistency.

    Generally lower latency is better. 300ms will be very different than 45-50ms but I am not sure if 20ms will make or break the deal for you. Are you outside the country? That is a pretty high ping time. You may want to investigate options geographically closer to your location to increase response time.

    Good luck!
    =>Admo.net Managed Hosting
    => Managed Hosting • Dedicated Servers • Colocation
    => Dark Fiber Access to 1102 Grand, Multiple Public Providers
    => Over •Sixteen• Years of Service

  10. #10
    Quote Originally Posted by AdmoNet View Post
    Hello,

    Write cache is a must for systems utilizing database engines like MySQL. Cache memory is always faster than disk. BBU keeps that cache persistent if power is lost. When the system is rebooted, cache is replayed to disk and you have consistency.

    Generally lower latency is better. 300ms will be very different than 45-50ms but I am not sure if 20ms will make or break the deal for you. Are you outside the country? That is a pretty high ping time. You may want to investigate options geographically closer to your location to increase response time.

    Good luck!
    I am in India but all my servers are in USA because there are not equally good hosts and connections here as yet. Thanks for letting me know that 20ms will not make much of a difference.

    From what I have heard here so far, Wiredtree server looks like a clear winner at the same cost.

    What about backups. With Raid1 already there, is an extra disk for backup not enough? Since I do not need to restore a single file do I really need R1soft instead of a backup disk?

  11. #11
    Join Date
    Jul 2008
    Posts
    39
    Quote Originally Posted by WiredTree Zac View Post
    No, actually we haven't had any network issues or outages in the last 2 months...
    I think he's referring to your extended downtime in Dec. and Jan., not to mention the DDoS attacks on Grove in Jan.

    I'm still trying to find the time to write a review of my migration from WiredTree to ServInt but I couldn't be happier. Their Managed Migration was simply magnificent.

  12. #12
    Join Date
    Oct 2003
    Location
    Chicago, IL
    Posts
    657
    Quote Originally Posted by DGRing View Post
    I think he's referring to your extended downtime in Dec. and Jan., not to mention the DDoS attacks on Grove in Jan.
    We did have a power issue which affected a small number of clients in a specific area of our datacenter 6-7 months ago, which was addressed immediately by our on-site staff. There were changes made to the power infrastructure in the datacenter to prevent that from happening again and everything has been 100% rock solid since then. The DDOS you are referring too was against our web site and Grove, not any client sites, so the impact was non-existent or minimal for nearly all of our clients.

    In any event, glad to hear you are happy again.
    Zac Cogswell / CEI
    Formerly known as WiredTree Zac

  13. #13
    Join Date
    Jul 2008
    Posts
    39
    Quote Originally Posted by WiredTree Zac View Post
    We did have a power issue which affected a small number of clients in a specific area of our datacenter 6-7 months ago which was addressed immediately by our on-site staff. There were changes made to the power infrastructure in the datacenter to prevent that from happening again and everything has been 100% rock solid since then. The DDOS you are referring too was against our web site and Grove, not any client sites, so the impact was non-existent or minimal for nearly all of our clients.
    I would beg to differ it was a one time thing with a "small number of clients":

    http://www.webhostingtalk.com/showthread.php?t=1005063

    http://www.webhostingtalk.com/showthread.php?t=1014815

    http://www.webhostingtalk.com/showthread.php?t=1021397

    Stats in the last thread from UnderHost:

    WiredTree from pingdom:

    February 2011
    Uptime (%): 94.55% Downtime (d, h, m) : 13 h, 38 m

    January 2011
    Uptime (%): 95.4% Downtime (d, h, m) : 1 d, 10 h, 14 m

    December 2010
    Uptime (%): 99.38% Downtime (d, h, m) : 48 m

    And I was confusing this DDOS:

    http://www.webhostingtalk.com/showthread.php?t=728315

    with this one in January.

    But it wasn't the downtime that motivated me to change vendors, it was the arbitrary changing of root passwords without notifying your customers:

    http://www.webhostingtalk.com/showthread.php?t=1019174

    That was unforgivable, imo.

    In any event, glad to hear you are happy again.
    Thanks, just didn't feel comfortable with your company anymore.

  14. #14
    Join Date
    Oct 2003
    Location
    Chicago, IL
    Posts
    657
    Quote Originally Posted by DGRing View Post
    I would beg to differ it was a one time thing with a "small number of clients":
    I'm not going to argue with you since you seem to have an axe to grind, however I have all the technical data in front of me (and I was there physically at the time) to say with 100% accuracy it was a small subset of clients we host. Sure, there are tons of posts in those threads (although many of the posts are by accounts created to complain in the thread and have never posted again...) but we have many, many times more clients than that.

    You are more than welcome to check our Hyperspin reports: http://www.hyperspin.com/en/ranking.php?type=3 where we have 99.997% uptime over 1020 days for our network.

    We're not immune to issues, nor are any other hosts, including your current one who has had a number of power/network issues as well since you have been with them. The test is how a host reacts to the issues, and we have done so properly and completely. That is why the only threads you find are half a year old.
    Zac Cogswell / CEI
    Formerly known as WiredTree Zac

  15. #15
    Join Date
    Aug 2002
    Location
    Seattle
    Posts
    5,525
    My problem with Hyperspin is the reliability of those reports. Last I checked, it was actually possible to setup monitors pointing to 127.0.0.1 and appear to have 100.00% uptime.

    Similarly, if you want to smear a competitor you can just point at random sites having poor uptime and mark them as being with that competitor.

  16. #16
    Can anyone post similar outage threads in the last one year and stats (from pingdom etc.) for Servint as well, so that a comparison can be made?

  17. #17
    Join Date
    Aug 2008
    Posts
    176
    Quote Originally Posted by aarkay777 View Post
    Can anyone post similar outage threads in the last one year and stats (from pingdom etc.) for Servint as well, so that a comparison can be made?
    http://www.webhostingtalk.com/search...rchid=32505361

    No one is immune from down time.

    For example Softlayer had a bad power outage several years ago. And Colo4 Dallis was knocked offline by a DDoS attack some time ago. Because one of their links was only GigE. So ya people go down its a fact it happens.

  18. #18
    Join Date
    Oct 2003
    Location
    Chicago, IL
    Posts
    657
    Quote Originally Posted by IRCCo Jeff View Post
    My problem with Hyperspin is the reliability of those reports. Last I checked, it was actually possible to setup monitors pointing to 127.0.0.1 and appear to have 100.00% uptime.

    Similarly, if you want to smear a competitor you can just point at random sites having poor uptime and mark them as being with that competitor.
    I don't know about the former as I've never tried, but you can contact Hyperspin about the latter.
    Zac Cogswell / CEI
    Formerly known as WiredTree Zac

  19. #19
    Quote Originally Posted by ShaunH View Post
    http://www.webhostingtalk.com/search...rchid=32505361

    No one is immune from down time.

    For example Softlayer had a bad power outage several years ago. And Colo4 Dallis was knocked offline by a DDoS attack some time ago. Because one of their links was only GigE. So ya people go down its a fact it happens.
    Clicking that link says "Sorry, no matches". Please check.

  20. #20
    Join Date
    Aug 2008
    Posts
    176
    Quote Originally Posted by aarkay777 View Post
    Clicking that link says "Sorry, no matches". Please check.
    Sorry I thought that would work.

    The point I was trying to make is, go to advanced search put in any providers name and search in the outages forum. Almost all hosts who have been around a while will have threads there. Even more so for those who are selling out of their own DC. Because they are the upstream to more users. So of course those folks are going to get more scrutiny.

    Whats funny is Softlayer has a lot of threads. But yet no one seems to find that a problem. Servint also has several threads as do Knowhost, and many many other hosts. But this doesn't make any of them them a bad host. Heck most people consider the hosts I listed to be very good.


    I think we all need to be realistic about up time. I would argue the quality of a host comes out during and after problems such as down time.

  21. #21
    Join Date
    Jul 2008
    Posts
    39
    Quote Originally Posted by WiredTree Zac View Post
    I'm not going to argue with you since you seem to have an axe to grind, however I have all the technical data in front of me (and I was there physically at the time) to say with 100% accuracy it was a small subset of clients we host. Sure, there are tons of posts in those threads (although many of the posts are by accounts created to complain in the thread and have never posted again...) but we have many, many times more clients than that.
    No axe to grind, just stating facts. But I find it difficult to believe that you would fly in the VP of CoreSite if it was such a "small subset of the clients we host".

    You are more than welcome to check our Hyperspin reports: http://www.hyperspin.com/en/ranking.php?type=3 where we have 99.997% uptime over 1020 days for our network.
    I don't have to check Hyperspin, I was a customer of your company during the period I was talking about. My problem was with the Dec., Jan., Feb. time frame, when this was your company's record:

    December 2010
    Uptime (%): 99.38% Downtime (d, h, m) : 48 m

    January 2011
    Uptime (%): 95.4% Downtime (d, h, m) : 1 d, 10 h, 14 m

    February 2011
    Uptime (%): 94.55% Downtime (d, h, m) : 13 h, 38 m

    We're not immune to issues, nor are any other hosts, including your current one who has had a number of power/network issues as well since you have been with them. The test is how a host reacts to the issues, and we have done so properly and completely. That is why the only threads you find are half a year old.
    I use monitor.us for my site and I've had 100% uptime since I've migrated to ServInt. Of course hosts are going to have problems, after all we're dealing with computers.

    However, I would disagree that your company reacted "properly and completely". Doing scheduled, non-emergent system/network maintenance during normal business hours in the U.S. is about as far from "best practice" as you can get. On the west coast of the U.S., people are still working at 7:00PM CST. And guess what happens when "routine maintenance" goes awry? My site, along with all of my client sites, go down when people are actually using them.

    And I won't even get into how you could allow a 3-party vendor to gain access to all usernames and passwords for the Grove system, which contains your customers credit card information. This was totally unrelated to the thread I mentioned above where root passwords were changed without notifying your customers, which you conveniently didn't respond to.

    We'll just have to agree to disagree on how well your company responded to three separate instances of downtime in a three month period. As I said in one of my posts during the outage:

    "How you and your company respond to these recent challenges will determine just how effective a manager you are... and whether I'll stay with you or not. I'm _so_ much in your corner and I wish you the very best of luck during this transition period."

    You quite obviously think your company responded perfectly and I, quite obviously, disagree.

    Which is why I and many, many others from that time period are no longer your customers.

  22. #22
    Join Date
    Oct 2003
    Location
    Chicago, IL
    Posts
    657
    Quote Originally Posted by DGRing View Post
    No axe to grind, just stating facts. But I find it difficult to believe that you would fly in the VP of CoreSite if it was such a "small subset of the clients we host".
    Why wouldn't they? They are just as concerned about uptime as we are - there was an issue they were taking seriously and they dealt with it that way. Pretty simple. We aren't the only tenants in the building, and others had issues, including internal CoreSite equipment, which you probably never gave thought to.

    I don't have to check Hyperspin, I was a customer of your company during the period I was talking about. My problem was with the Dec., Jan., Feb. time frame, when this was your company's record:
    That may have been your individual server or someone else's you dug up are posting to keep grinding your axe, but that wasn't our facilities uptime nor our networks (otherwise that would be reflected at Hyperspin). Since you will never believe me, I'm not going to argue with you.

    However, I would disagree that your company reacted "properly and completely". Doing scheduled, non-emergent system/network maintenance during normal business hours in the U.S. is about as far from "best practice" as you can get. On the west coast of the U.S., people are still working at 7:00PM CST. And guess what happens when "routine maintenance" goes awry? My site, along with all of my client sites, go down when people are actually using them.
    We always schedule maintenance on off-hours (midnight or after CST) unless there is emergency maintenance, which due to the nature of being an "emergency" would need to be performed at another time.

    And I won't even get into how you could allow a 3-party vendor to gain access to all usernames and passwords for the Grove system, which contains your customers credit card information. This was totally unrelated to the thread I mentioned above where root passwords were changed without notifying your customers, which you conveniently didn't respond to.
    This is not true. Please refrain from making further libelous statements.
    Zac Cogswell / CEI
    Formerly known as WiredTree Zac

  23. #23
    Join Date
    Nov 2009
    Location
    Cincinnati
    Posts
    1,585
    Quote Originally Posted by DGRing View Post
    I would beg to differ it was a one time thing with a "small number of clients":

    http://www.webhostingtalk.com/showthread.php?t=1005063

    http://www.webhostingtalk.com/showthread.php?t=1014815

    http://www.webhostingtalk.com/showthread.php?t=1021397

    Stats in the last thread from UnderHost:

    WiredTree from pingdom:

    February 2011
    Uptime (%): 94.55% Downtime (d, h, m) : 13 h, 38 m

    January 2011
    Uptime (%): 95.4% Downtime (d, h, m) : 1 d, 10 h, 14 m

    December 2010
    Uptime (%): 99.38% Downtime (d, h, m) : 48 m

    And I was confusing this DDOS:

    http://www.webhostingtalk.com/showthread.php?t=728315

    with this one in January.

    But it wasn't the downtime that motivated me to change vendors, it was the arbitrary changing of root passwords without notifying your customers:

    http://www.webhostingtalk.com/showthread.php?t=1019174

    That was unforgivable, imo.



    Thanks, just didn't feel comfortable with your company anymore.
    Knew I wasn't the only one who had noticed it :-)

    Hey WT, what data center do you use and do you own your network or are you leasing?
    'Ripcord'ing is the only way!

  24. #24
    Join Date
    Oct 2003
    Location
    Chicago, IL
    Posts
    657
    Quote Originally Posted by Visbits View Post
    Hey WT, what data center do you use and do you own your network or are you leasing?
    We have multiple footprints in 427 S LaSalle (CoreSite) in downtown Chicago. We own and operate our own network - AS19066 with 80Gbps connectivity from 4 tier-1 (L3, GBLX, Savvis, NTT) providers running primarily on the new Brocade MLXe platform. The network upgrades were recently completed, so our web site is a bit of of date and is scheduled to be updated.
    Zac Cogswell / CEI
    Formerly known as WiredTree Zac

Similar Threads

  1. Replies: 15
    Last Post: 03-21-2010, 07:18 PM
  2. ServInt / Knownhost / Wiredtree
    By tender_heart in forum VPS Hosting
    Replies: 24
    Last Post: 12-21-2009, 02:40 AM
  3. Wiredtree OR Servint?
    By meerk in forum VPS Hosting
    Replies: 26
    Last Post: 10-19-2009, 04:21 PM
  4. WiredTree --> Servint?
    By reloaded in forum VPS Hosting
    Replies: 43
    Last Post: 04-10-2009, 02:39 PM
  5. ServInt vs WiredTree
    By arcturus in forum VPS Hosting
    Replies: 4
    Last Post: 08-28-2008, 10:26 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •