Results 1 to 22 of 22
  1. #1

    [ask] Should i upgrade my ram?

    I'm in a vps with 512 MB ram now, my visitors per day is 10K and from top command, i just use about 200 MB. But, my vps is always got downtime everyday about 30 minutes. The CS said it is because my ram, the top command is not representing the real ram and recommend me to upgrade my ram to 768 MB.

    Should i upgrade my ram?

  2. #2
    Join Date
    Sep 2006
    Location
    Dallas, TX
    Posts
    333
    I would not think downtime is related to RAM usage. Your web site would be slow if your machine is swapping, but it should not be down.

    How are you determining downtime? If you are using a web request, that would indicate either downtime or non-responsiveness. If you are using ping or some other monitoring, that would be a different problem.

  3. #3
    Join Date
    Oct 2009
    Location
    Nottingham, UK
    Posts
    57
    upgrading your ram isn't always the answer - it's most likely the additional stuff that comes with upgrading to a higher package which is why they want you to upgrade. From a hosts point of view (depending on the type of virtualization software being used) they can specify number of input/output requests allowed at any one time along with maximum disk usage and a whole bunch of other stuff.

    What i'm trying to say in a very round about way is most likely it's not the ram your using up but another limit they have in place for your account. Many hosting providers overload their VPS setups and so place quite strict limits in place on things so whilst you may have an allocation of 512MB of ram you'd never be able to use it as so many other resource limits are in place you'll hit those long before you hit any other which might well be why your VPS keeps going down.

    Of course it may also be their hardware node is to blame so the first thing is to identify when your vps goes down has the whole hardware node gone down. If the answer is no then you're hitting another limit and it might well be the effect of upgrading increases these limits.

    I have to say though if you're getting 10K visits per day then I would think about upgrading to a dedicated server as whilst the cost is higher you wouldn't have these kind of resource issues...
    SWORDFISH HOSTING - UK Based Web Hosting - 24*7 Support
    Shared Hosting | VPS | Dedicated Servers | Plesk 9 Control Panel
    Web Hosting For Everyone

  4. #4
    Quote Originally Posted by NeilAgg View Post
    I would not think downtime is related to RAM usage. Your web site would be slow if your machine is swapping, but it should not be down.

    How are you determining downtime? If you are using a web request, that would indicate either downtime or non-responsiveness. If you are using ping or some other monitoring, that would be a different problem.
    I'm using pingdom 1 minute monitor from different country. So, i will know the website is not only downtime from my computer.

    Quote Originally Posted by SwordfishHostingLtd View Post
    upgrading your ram isn't always the answer - it's most likely the additional stuff that comes with upgrading to a higher package which is why they want you to upgrade. From a hosts point of view (depending on the type of virtualization software being used) they can specify number of input/output requests allowed at any one time along with maximum disk usage and a whole bunch of other stuff.

    What i'm trying to say in a very round about way is most likely it's not the ram your using up but another limit they have in place for your account. Many hosting providers overload their VPS setups and so place quite strict limits in place on things so whilst you may have an allocation of 512MB of ram you'd never be able to use it as so many other resource limits are in place you'll hit those long before you hit any other which might well be why your VPS keeps going down.

    Of course it may also be their hardware node is to blame so the first thing is to identify when your vps goes down has the whole hardware node gone down. If the answer is no then you're hitting another limit and it might well be the effect of upgrading increases these limits.

    I have to say though if you're getting 10K visits per day then I would think about upgrading to a dedicated server as whilst the cost is higher you wouldn't have these kind of resource issues...
    Thank you very much, for your comment.

    I don't know what other limit they use, but i think the issue come from apache. If i use "uptime" command in ssh, the uptime server said more than 30 days, but when i see apache status in cpanel, i just can see about 6 hours uptime in apache.
    Is it something like this in apache is normal?
    Last edited by nee4; 11-09-2009 at 12:08 PM.

  5. #5
    Join Date
    Apr 2007
    Posts
    3,531
    As other people have said RAM shouldn't cause downtime unless you have so little of it left programs just crash, even then swapping starts so they just go slower, alot slower!

    Ask your host for some kind of proof that your using all your ram, doesn't your control panel show you this?
    BotWars.io - Code the AI of your Battle Bot!

  6. #6
    Join Date
    Oct 2009
    Location
    Nottingham, UK
    Posts
    57
    if apache is saying it's been up then it's more likely that it's a local thing to you related to the number of hits you get.

    The best thing to do is to check the httpd logs for the time you know the server was offline. I've had quite a few problem in the past related to the number of child processes apache was allowed to spawn and I had to increase that but the httpd logs clearly told me that so i'd start there.
    SWORDFISH HOSTING - UK Based Web Hosting - 24*7 Support
    Shared Hosting | VPS | Dedicated Servers | Plesk 9 Control Panel
    Web Hosting For Everyone

  7. #7
    Join Date
    Feb 2004
    Location
    Sacramento CA
    Posts
    3,513
    I would ask your host exactly how ram is causing your downtime and ask for clear evidence. I can see some processes being killed or stopped from spawning but bringing down your whole website? If you don't get an answer that makes sense I would start looking for a new host.

  8. #8
    I would do what Swordfish was saying check your apache logs and system logs, see how what the logs say match up to what your host is telling you. If they are blatantly wrong and just trying to sell you on an upgraded system ask to talk to someone higher up and see if they could be more helpful or look into another host.

  9. #9
    Thanks all, i'll check my httpd log first. I will post here again if there is something wrong with my httpd log.

    Anyway, where is the directory of httpd logs? Is there any other logs that can help me fix this issue?
    Last edited by nee4; 11-10-2009 at 03:32 AM.

  10. #10
    Join Date
    Oct 2009
    Location
    Nottingham, UK
    Posts
    57
    changes depending on the system but typically /var/log/httpd/access_log

    if you are unfamiliar with grepping through the file to find errors then the easiest option is to copy that file (note copy not move ) to a directory you can ftp to and download it. You'll be able to open it in notepad and see the timestamps for periods where you're server was offline.
    SWORDFISH HOSTING - UK Based Web Hosting - 24*7 Support
    Shared Hosting | VPS | Dedicated Servers | Plesk 9 Control Panel
    Web Hosting For Everyone

  11. #11
    I would upgrade to atleast 1 gig tbh. You have low ram.

  12. #12
    Join Date
    Sep 2009
    Posts
    1,976
    You can ask your hosting provider to upgrade your RAM memory for a couple of days, just to test. Once you are sure that the server is working without a problem, after that RAM memory upgrade than you can pay them. I'm sure that if they are sure that the RAM memory is the problem they would not mind to upgrade your account for a couple of days.

  13. #13
    Join Date
    Nov 2009
    Posts
    514
    When you say downtime, is the vps unresponsive or just apache?
    www.VMhosts.co.uk - "Cloud hosting within reach"
    UK VMware Cloud IaaS, 24x7x365 Support, Evault Cloud Backup & DRaaS
    sales@vmhosts.co.uk 020 3397 1233

  14. #14
    Join Date
    Sep 2008
    Location
    Dallas, TX
    Posts
    4,568
    Quote Originally Posted by nee4 View Post
    I'm in a vps with 512 MB ram now, my visitors per day is 10K and from top command, i just use about 200 MB. But, my vps is always got downtime everyday about 30 minutes. The CS said it is because my ram, the top command is not representing the real ram and recommend me to upgrade my ram to 768 MB.

    Should i upgrade my ram?
    Are you using a xen VPS?


    You can ask your provider if they'll upgrade you for a day to a higher amount of RAM to see if it resolves the problem.
    Jacob Wall

  15. #15
    Join Date
    Feb 2005
    Location
    Australia
    Posts
    5,849
    Quote Originally Posted by nee4 View Post
    I'm in a vps with 512 MB ram now, my visitors per day is 10K and from top command, i just use about 200 MB.
    Based on that I'd speculate that it's Virtuozzo with SLM, but perhaps the OP can confirm. If it's openVZ the user_beancounters file will usually tell you everything you need to know about memory usage.

    But the first thing to establish, as already stated above, is exactly what's meant by "downtime". Without more details it could mean, eg. node down, VPS down, services down (MySQL / httpd), script errors, dns lookup failures etc.
    Chris

    "Some problems are so complex that you have to be highly intelligent and well informed just to be undecided about them." - Laurence J. Peter

  16. #16
    Quote Originally Posted by VMhosts View Post
    When you say downtime, is the vps unresponsive or just apache?
    Just the apache, i'm sure of it. Because while the site down, i'm still logging in my server via ssh.

    Quote Originally Posted by Jacob Wall View Post
    Are you using a xen VPS?

    You can ask your provider if they'll upgrade you for a day to a higher amount of RAM to see if it resolves the problem.
    I ask them and said Virtuozzo with SLM, anyway, what is the different between them?

    Quote Originally Posted by foobic View Post
    Based on that I'd speculate that it's Virtuozzo with SLM, but perhaps the OP can confirm. If it's openVZ the user_beancounters file will usually tell you everything you need to know about memory usage.

    But the first thing to establish, as already stated above, is exactly what's meant by "downtime". Without more details it could mean, eg. node down, VPS down, services down (MySQL / httpd), script errors, dns lookup failures etc.
    As i say, just the apache down, i'm sure of it. Because while the site down, i'm still logging in and editing my server via ssh. I check the server uptime via ssh and it say about 30 days, while pingdom report is down about once in 2 days.

  17. #17
    Join Date
    Feb 2005
    Location
    Australia
    Posts
    5,849
    Quote Originally Posted by nee4 View Post
    I ask them and said Virtuozzo with SLM, anyway, what is the different between them?
    SLM means not having to worry about the weirdness that is openVZ's beancounters...
    As i say, just the apache down, i'm sure of it. Because while the site down, i'm still logging in and editing my server via ssh. I check the server uptime via ssh and it say about 30 days, while pingdom report is down about once in 2 days.
    That confirms that the host node and your VPS are still up. It doesn't distinguish between apache down vs script errors vs dns lookup fails. What errors is pingdom reporting?

    For more information about services down (and to restart them automatically) you might want to try SIM.
    Chris

    "Some problems are so complex that you have to be highly intelligent and well informed just to be undecided about them." - Laurence J. Peter

  18. #18
    Join Date
    Sep 2008
    Location
    Dallas, TX
    Posts
    4,568
    Quote Originally Posted by nee4 View Post
    Just the apache, i'm sure of it. Because while the site down, i'm still logging in my server via ssh.


    I ask them and said Virtuozzo with SLM, anyway, what is the different between them?



    As i say, just the apache down, i'm sure of it. Because while the site down, i'm still logging in and editing my server via ssh. I check the server uptime via ssh and it say about 30 days, while pingdom report is down about once in 2 days.
    Login VZCP - https://vpsip:4643/

    Go to "Resource Alerts" and look at what it says. Attached is what a VPS's Resource Alert looks like, it might or might not look like this please attach yours.
    Attached Thumbnails Attached Thumbnails Picture 9.png  
    Jacob Wall

  19. #19
    Sorry, long time doesn't give reply. Yesterday it going downtime again, but i take screenshot from the control panel

    You can see the "load average" when my site down, it going red, but the memory i use still 28%, what is this mean?
    Attached Thumbnails Attached Thumbnails Downtime.jpg   Uptime.jpg  

  20. #20
    Join Date
    Feb 2005
    Location
    Australia
    Posts
    5,849
    If that's as bad as it gets you're not running out of memory. The slightly elevated load could indicate cpu or disk i/o as a problem but I wouldn't think that should be high enough to cause timeouts.

    Are you sure it's not a dns lookup failure? What did the pingdom report say exactly?
    Chris

    "Some problems are so complex that you have to be highly intelligent and well informed just to be undecided about them." - Laurence J. Peter

  21. #21
    Join Date
    Oct 2009
    Location
    South Padre Island,Texas
    Posts
    16

    *

    Quote Originally Posted by nee4 View Post
    I'm in a vps with 512 MB ram now, my visitors per day is 10K and from top command, i just use about 200 MB. But, my vps is always got downtime everyday about 30 minutes. The CS said it is because my ram, the top command is not representing the real ram and recommend me to upgrade my ram to 768 MB.

    Should i upgrade my ram?
    your sshot looks like 9gb not 200 mb try to turn off your email, all 3.

    i think you have been hacked. i had 100 spam emails a minute.

    the spam guard turned on just did not stop mine.



    Warren Buffet on gold:

    It gets dug out of the ground in Africa or someplace. Then we melt it down, dig another hole, bury it again and pay people to stand around guarding it.

  22. #22
    Quote Originally Posted by foobic View Post
    If that's as bad as it gets you're not running out of memory. The slightly elevated load could indicate cpu or disk i/o as a problem but I wouldn't think that should be high enough to cause timeouts.

    Are you sure it's not a dns lookup failure? What did the pingdom report say exactly?
    The pingdom error said:
    Connection refused
    HTTP CRITICAL - Unable to open TCP socket

    Quote Originally Posted by kemosabetx View Post
    your sshot looks like 9gb not 200 mb try to turn off your email, all 3.

    i think you have been hacked. i had 100 spam emails a minute.

    the spam guard turned on just did not stop mine.



    Warren Buffet on gold:

    It gets dug out of the ground in Africa or someplace. Then we melt it down, dig another hole, bury it again and pay people to stand around guarding it.
    Can you give me more details about this, what do you mean by
    "sshot looks like 9gb not 200 mb"

    Thanks, for the reply

Similar Threads

  1. Replies: 4
    Last Post: 11-27-2009, 04:49 AM
  2. SOFTLAYER - Double HDD/RAM continues, or Double Bandwidth!
    By maryhall in forum Dedicated Hosting Offers
    Replies: 1
    Last Post: 11-09-2009, 04:28 PM
  3. Replies: 0
    Last Post: 11-08-2009, 09:59 PM
  4. Core Hoster - GreenCore Server Upgrade - Ram Upgrade - (Scheduled)
    By VbCrazy in forum Providers and Network Outages and Updates
    Replies: 0
    Last Post: 04-01-2009, 08:28 PM
  5. Dual Opteron Blowout!! Free CPU Upgrade or 15k SCSI Upgrade!
    By Jason K in forum Dedicated Hosting Offers
    Replies: 6
    Last Post: 11-02-2006, 02:42 PM

Posting Permissions

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