Page 2 of 4 FirstFirst 1234 LastLast
Results 26 to 50 of 83
  1. #26
    Join Date
    Jul 2001
    Location
    Sydney NSW
    Posts
    82
    I have 2 boxes with rackshack (never any problems) for 3 or 4 years.

    I have 1 box with servermatrix for 3 months (HEAPS of problems due to Enterprise). Need to send reboot request almost daily.

    I have 1 box with managed.com for 3 months (no problems). Needed reboot once - was done in 30 seconds! ???

    I am about to order second one with managed. Servermatrix probably has faster network, but the managed boxes have MUCH better hardware specs.

  2. #27
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    In both cases

    are they making backup to my harddrive???

    or do I have to buy/rent another harddrive for backup??

  3. #28
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    another question is:

    can i sell accounts?

    I mean I want to be a reseller using one server.

    Regards

  4. #29
    Join Date
    Feb 2004
    Location
    Sofia
    Posts
    1,354
    Originally posted by WebDork
    I have 1 box with servermatrix for 3 months (HEAPS of problems due to Enterprise). Need to send reboot request almost daily.
    I am afraid this is the bare SM truth. I even think of starting a cron job for automatically restarting Apache every two hours to prevent CPU overload and memory incremental use ending beyond the 95% limit. Changed to the last RHE kernel, bought some extra RAM, and nothing. Instead of doing my business I am changing httpd.conf every four hours, hoping that this time the server will be up for more than 10 hours.

    The whole mess started on March 12th...
    Last edited by Imago; 04-02-2004 at 06:39 AM.

  5. #30
    Join Date
    Feb 2004
    Location
    Sofia
    Posts
    1,354
    mohdjumah,

    You certainly can all these things. But better use external drives for backup.

  6. #31
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    Imago,
    Thanks for the info. you provided and your quick reply.

    By external drives (do u mean another server?)

    anybody knows how long is the drive backup should be? if I have 80/120 GB for doning full drive backup .tar.gz high compressed?

    accually I want to rent a server, that's why I'm asking alot of question and I hate sending Questins to support coz they are useless/take time to reply and I'm sure they will give a link to their FAQ which I HATE IT.

    Regards

  7. #32
    Join Date
    Feb 2004
    Location
    Sofia
    Posts
    1,354
    Yes, external means on a different server and possibly in different geographic location. There are some very inexpensive backup servers. If you use the full account compression as provided by cPanel, you will need 10 to 20 Gigs for the drives you specified.

  8. #33
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    I appreciate your experience.

    Thanks a lot.

  9. #34
    Join Date
    Feb 2004
    Posts
    208
    Originally posted by WebDork
    I have 2 boxes with rackshack (never any problems) for 3 or 4 years.

    I have 1 box with servermatrix for 3 months (HEAPS of problems due to Enterprise). Need to send reboot request almost daily.

    I have 1 box with managed.com for 3 months (no problems). Needed reboot once - was done in 30 seconds! ???

    I am about to order second one with managed. Servermatrix probably has faster network, but the managed boxes have MUCH better hardware specs.
    I guess your rackshack servers are older version of RH7.x and managed servers are installed with RH9.

    I have servers from EV1 and managed, which have no problem so far, but I notice Plesk is much more stable then Ensim Pro. Never use CP before.

    Your SM server down frequently due to RHE. I thought RHE is a better O/S and I see a lot of DC now offer RHE by default, like EV1 installedl RHE by default.

    I even ask managed why they don't offer RHE like other DC. I guess before RHE stabilize we better stick with the older RH.

  10. #35
    Join Date
    Jul 2001
    Location
    Sydney NSW
    Posts
    82
    Yep. I am going to stay with RH9 on my servers for at least another few builds of Enterprise. Plesk 7 is pretty good in my opinion. Seems very stable, and very easy to use.

  11. #36
    Join Date
    Jul 2001
    Location
    Sydney NSW
    Posts
    82
    Oh - and I did indeed order another Managed.com server. They just cant be beat on price for the hardware you get. Network seems solid as well.

  12. #37
    Join Date
    Feb 2004
    Posts
    208
    Originally posted by webrats
    10mbps sucks, they just do that so they can rip people down the line.
    i like a network card that can burst over 10. im sure you customers wont be to happy if they slow down to a crawl because of crappy hardware.
    but what do i know i do about 30gb a day. Im one of those people who use the server they buy
    I have reservation on 10mbps also. We don't know how many servers shared the same network, so the burstable bandwidth is very much limited. If you are on the same network with game servers, that's it those servers will eat up all the bandwidth.

  13. #38
    Join Date
    Jul 2001
    Location
    Sydney NSW
    Posts
    82
    Just for interest - does anyone know if the 2.6 hyperthreading CPU will perform much better than a 2.8 ?

  14. #39
    Join Date
    Feb 2004
    Posts
    208
    Originally posted by WebDork
    Oh - and I did indeed order another Managed.com server. They just cant be beat on price for the hardware you get. Network seems solid as well.
    I have 2 managed servers already. Transferring my clients over from EV1. But I am doing this cautiously just in case there is a relapse of pre-XO problems. If another month managed maintain such solid stability I might give up EV1 altogether.

  15. #40
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    Originally posted by WebDork
    Just for interest - does anyone know if the 2.6 hyperthreading CPU will perform much better than a 2.8 ?
    I used Intel P4 3.06G HT Processor With 1024 MB RAM. But even the preformance of 2.4G was better.

  16. #41
    10mbps isn't shared, it's just your port. Yes, you all share the backbone connections, but no, you don't share 10mbps with other servers.

  17. #42
    Join Date
    Feb 2004
    Posts
    208
    Originally posted by mohdjumah
    I used Intel P4 3.06G HT Processor With 1024 MB RAM. But even the preformance of 2.4G was better.
    You mean P4 3.06G HT is NOT faster than P4 2.4G?

  18. #43
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    Originally posted by mcraft
    You mean P4 3.06G HT is NOT faster than P4 2.4G?
    Yes, The preformance test we had approved that 2.4G/2.8G is better at PREFORMANCE than 3.06G. However, 3.06 is faster in resulting the instructions, and freezed a lot. Maybe Intel needs to improve the next generation of HT's.

  19. #44
    Join Date
    Feb 2004
    Location
    Sofia
    Posts
    1,354
    I have the same feeling about 3.06...

  20. #45
    Join Date
    Feb 2004
    Posts
    208
    Phew! I almost ordered a P4 3G HT, but the price hold me back. Ok I will stick with non HT for the time being. EV1 is charging P4 3G HT same price as dual xeon servers, sorry for those who think HT is better.

  21. #46
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    BTW. In UAE/Dubai 3.06G Vs. 2.8G was around $250. But nowadays its not more than $50 difference.

  22. #47
    Join Date
    Feb 2004
    Posts
    208
    On second thought must the O/S be made for HT in order to exploit HT fully? Is it possible that the existing O/S still not HT ready yet?

  23. #48
    Join Date
    Apr 2004
    Location
    UAE
    Posts
    62
    Originally posted by mcraft
    On second thought must the O/S be made for HT in order to exploit HT fully? Is it possible that the existing O/S still not HT ready yet?

    OS must not have to deal with Processor speed technology. Processor is only processing commands instructions whatever is the software. Also, no tweaking officially came up from Intel for a certain OS to perform the HT Processors technology.
    Last edited by yawsh; 04-02-2004 at 10:07 AM.

  24. #49
    Join Date
    Feb 2004
    Location
    Your Screen
    Posts
    3,999
    Hi

    This thread has gone in a variety of directions so I will do what I can to bring it together and address the issues presented with my own experiences.

    I have a 2.8 GHz P4, 2 GB RAM, running RHES 3.0, at ServerMatrix. This is a non-HT box.

    I have identical servers running at two other data centers. These identical servers are all running RH 7.3.

    A couple of months ago I was faced with the choice of whether to upgrade my SM box to a dual Xeon (which I desperately wanted to do) or stick with the trustly P4. I spoke at length with SM's sales team about this, and also spent a couple of days reading through every HT/RHES thread I could find. At that time the only prescribed fix was tentative at best, and involved deploying the dual Xeon under an old/insecure kernel. That is absolutely not an option for us; we need to have our boxes secure. Their alternative recommendation was to deploy the box under RH 9 and then reload the box under RHES at a later date (which she said they would do for free, but I didn't want my clients to be saddled with downtime).

    Sales made it very clear the dual Xeons were not a stable platform as I wished to have one deployed. Hence I was left with the choice to take a chance on a stylin' dual Xeon with a big "Unknown" in terms of stability, or stick with the trusty P4's which have served us so well thus far. We decided to stick with reliability, and continue to deploy on P4's for now.

    Our next box will be at a different provider than SM, and I do plan to get a dual Xeon and test it extensively before making it production. It does seem that it is only the SM and EV1 dual Xeons that are having issues. I can't help but postulate that they are both using the same provider/parts and something with the hardware configuration is not playing nicely. It is obviously an as-yet unidentified bug with this particular config. Other data centers do not seem to be running into this issue. But, short of popping the case on a SM/EV1 box, vs. another data center's box, we may likely never know the real root cause here -- who has insider access to supplier names, model numbers and specifics for multiple data centers? Most providers keep that stuff tightly under wraps.

    Our P4's are rock solid. We don't have a single complaint.



    As far as connection sizes... capping servers at 10 Mbps is a common practice in the industry. The standard config for dedicated servers at both VenturesOnline and Dinix, for example, is also 10 Mbps NICs. The reasoning offered by all providers I have contacted/worked with on this, is exactly as explained earlier, that it protects you from a world of hurt if your box is DOS'd. The dedicated client is responsible for b/w charges if there is a DDoS. Those charges add up really fast on a 100 Mbps NIC, as opposed to a 10 Mbps.

    That said, is a 10 Mbps box the product for everyone? Well of course not and it makes absolutely no sense for those who clearly need a 100 Mbps NIC for their uses to come here and prescribe that anything less is bad business. That is simply not true. Everyone here has different needs and uses; that is why servers are made in different configurations, and why hosting companies do offer different options. The poster of the thread did not indicate what kind of data he is pushing. For someone to presume in either direction -- 10 Mbps or 100 Mbps -- is inappropriate. Obviously these are two totally different clienteles/kinds of need. I am sure the poster can look at the companies' sites and decide on the best price based on the published NIC prices.


    All this said... my experience with SM support... is outstanding. Their techs are well-trained and very knowledgeable, and really give a sh*t about their job (in my experience). This past week I worked with Bo and Zach on a server over the course of 18 hours and they were absolutely terrific. They updated the tickets when they could, but as someone who works directly with on-site techs myself, I understand that when a tech is tied up with a box, they do not have time/opportunity to break and take extra time to hold my hand in a ticket. I absolutely want them focusing on the task at hand, interruptions lead to mistakes. But I must clarify that in this past week...
    • My reboot request had a tech dispatched to it immediately. When I called 6 minutes later to share my observations, in the hopes of catching the tech who'd be going to the floor, I found out the tech was already at the server working on it.
    • The reboot tech had the server re-configured to deal with a compromise, and had the box back up and online, very quickly. It was especially quick considering the troubleshooting and custom configs he had to do to get the box back up. We were very impressed. He also provided us with key information necessary for diagnosing the issue with the box -- no wild goose chases here. He gave us what we needed to know, immediately.
    • After collecting data and doing our own diagnostics, we put in for a full reload. I followed that up with a phone call to ask what kind of time-frame we might be looking at with the reload. This is when I connected with Bo. While on the phone I explained the series of events, symptoms, our diagnostics, and inquired what their schedule looked like at the time? Just so I would have an idea to tell clients and do some planning on our end. Naturally he couldn't make any promises other than quote their 24-hour reload SLA, which was totally cool. I explained what timing would work best for us, but that we'd take what we could get.
    • The box went down for the reload 45 minutes later.
    • Unfortunately SM's imaging system was on the fritz, or our new drive wasn't feeling cooperative, so the OS needed to be reloaded by hand. This was followed by hardware configuration issues with the box. Again, I am very familiar with on-site issues, I know the issues reloads encounter and nothing here was out of line at all. When it was all said and done, the box was pingable within 5.5 hours, and fully online by the 6 hour mark.


    The SM techs provided outstanding follow-up support as we requested further information about the box and configs, responding to our tickets within a half-hour each time and logging in to make sure things were working as they should.

    It took a bit longer than we all wanted for the box to go live, but the box was 100% operational and is completely stable. That's the kind of service I want -- I don't want a rush job where it's slapped together with our fingers crossed. They took the time to iron out the issues the right way, and as a result they put up a happy box. That makes for happy customers.


    Throughout the process we did call just a couple of times to ask about how things were going. They weren't pushy calls, just gentle "how's it going?"'s with a smile and plenty of encouragement and thank-you's. The one time I asked for Bo and he wasn't available, well this is the call when we hooked up with Zach in the process and I just explained the scenario and that I was simply curious how it was going and making sure everything was OK. He offered several times to have Bo call me back, and I told him that wasn't necessary if everything was OK... that we weren't pushing at ALL, just checking in to see how it was going. In return I rec'd two updates to our ticket within a half-hour !! -- and the guys were very friendly and reassuring. They worked really hard on this system and they did a great job getting it done.

    Nobody likes paying for a reload, LOL, but it was not money wasted. They did a super job and it was quality work. They get a two thumbs-up from here, on all accounts:

    ** SPEED
    ** QUALITY OF WORK
    ** TECHNICAL SKILL
    ** TIMING AND CONTENT OF TICKET UPDATES
    ** HELPFULNESS OF PHONE SUPPORT


    SM has met our needs and has taken care of use admirably to date.

    Bailey
    Last edited by bithost(NET); 04-03-2004 at 12:17 AM.
    Let's Connect on Twitter! @thatsmsgeek2u || Fighting mediocrity one thread at a time.

  25. #50
    Join Date
    Apr 2004
    Location
    Chicago
    Posts
    46

    Re: ServerMatrix Or Managed.com?

    I had a peek at ServerMatrix.

    Look at the full dedicated server list. From the third one down on forward, they list the video memory for each configuration. Huh?

    "No way in hell am I running a web server without a Radeon 9000 or better!"

Page 2 of 4 FirstFirst 1234 LastLast

Posting Permissions

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