Howdy, Stranger!

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

xprotoceptxprotocept Member
edited August 2011 in General


LEB loads really slow for me, does everyone has this?


  • FranciscoFrancisco Top Host, Host Rep, Veteran

    I've been getting reports of cloudflare being dead on #frantech for the past little while. It seems they've been likely drawing in a lot of ddos pounded websites.


  • Clearly ok now! :)

  • Very slow for me currently. And, no I'm not on the 5 minute for a page to load network. :)

  • I found the performance on CloudFlare really depends on where you are connecting from, since it's geographical distributed reverse proxy. Sometimes I'll get reverse proxy error (CloudFlare reporting site down), connect via a different proxy and it's fine.

    As of the site itself, I don't think it has gone down over the last 7 days.

    $ uptime
     23:14:49 up 39 days, 20:01,  1 user,  load average: 0.46, 0.38, 0.25

    Yeah I know the load is a little bit high. I guess I need a little bit more memory :P

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    LowEndAdmin said: I guess I need a little bit more memory :P

    You heard it here folks!


  • Hehehe. I have been trading memory for performance all along -- no MySQL query caching, minimal key-buffers, no PHP opcode caching, no page-level caching etc. However sometimes when you need a little bit of extra speed, extra memory can become handy when you use it wisely.

  • I just spent about 12 minutes trying to get to this site. (7:45pm Tuesday US eastern time) Wanderingwifi kept telling me over and over again that it was down and forwarding me unto OpenDNS' lookup.

  • dmmcintyre3dmmcintyre3 Member
    edited August 2011

    How are you using a full 80mb with "no MySQL query caching, minimal key-buffers, no PHP opcode caching"? My 128mb KVM box with 2-3 PHP processes, query cache and largish key-buffer, NSD, pptpd, APC and some more stuff uses around 60mb when there is 3 php processes and the backup script is running. It's at 46mb with 2 php processes and the backup script not running right now.

  • Do note that 128MB has 60% more memory than 80MB. Not to mention that there is fixed overhead like kernel which would take up a larger percentage on 80MB. My OS-level disk cache is also at around 25MB, which would be greatly reduced when backup is running (rdiff-backup, which can use up to 12MB + extra shell/ssh processes).

    WordPress 3+ has also been quite heavy

    My 128mb KVM box with 2-3 PHP processes, query cache and largish key-buffer

    Mind posting your my.cnf and the result of ps aux to show what you meant by "largish key-buffer"? Default Debian install uses 16MB which I consider pretty small for this site...

  • dmmcintyre3dmmcintyre3 Member
    edited August 2011

    It's always below 65mb ram used except for when the backup script locked up and spawned 5 rsync processes at once.


    default-storage-engine = myisam
    key_buffer = 8M
    query_cache_size = 12M
    query_cache_limit = 512k
    tmp_table_size = 4M
    max_heap_table_size = 4M

    ps aux:

    I do need to clean some stuff up, like thttpd is not needed anymore. (just need to convert one .cgi to php)

  • Yeah. Just adding up the RSS for MySQL, PHP-FPM and Nginx it would be 60MB already. For MyISAM tables, having a larger key_buffer is usually a good idea (rather than a larger query cache, but then it really depends on the application). And

    • acpid is probably not needed
    • you probably only need 1 worker process for nginx (rather than 2)
    • is your /bin/sh bash or dash? Process 2057 seems to be a bit high at 1244kb

    I am also running Postfix on this low end box which is quite a bit heavier than exim4-light. Used for something else before but not anymore -- maybe time to roll back to exim4 again.

  • I just want to know what your big ass bloated VMware box is doing? ;)

  • FranciscoFrancisco Top Host, Host Rep, Veteran

    bitcoin, duh :D

    Thanked by 1Brandon_eNetSouth
  • +1

    Truly I was just trying to give him a poke or two and see what the comment would be. I never really see much out of it anymore. With the other forum software it had some crazy issue that would max out at 4Ghz CPU usage and just sit there.

  • That big ass bloated VMware box is serving this very comment.

    /me press [F5] a few times to make it a little bit busy.

    Actually the DDoS also killed quite a bit of traffic and now LowEndTalk is just hovering at around 100k page views/month that in theory does not need a big ass box (a 512MB memory Debian 6 btw, for those who are wondering).

  •              total       used       free     shared    buffers     cached
    Mem:           122         88         33          0          1         36
    -/+ buffers/cache:         50         71
    Swap:          258          0        258

    Also, it appears I'm getting a small GET flood attack from 1 IP right now, need to fix that.

  • You're still running without a cache, aren't you? I seem to recall you mentioning that a few times.

  • InfinityInfinity Member, Host Rep

    I don't think he is running a cache.

  • @drmike -- me or @dmmcintyre3?

    Most WordPress caching solutions require buffering the page result in memory first, which actually make it slower...

  • You, LEA. We run the wp-super-cache plugin in half mode for our (What we still call) wpmu installs. Just the common details in memory. It's a help.

    Saving the static page would be a minus but I think more so because of the number of comments that you get. All those regenerations...

  • drmikedrmike Member
    edited August 2011

    LEB appears to be having issues. Got it to load once and now the local proxy is saying that it can't find the site.

    edit: Spoke too soon. Comes up now.

Sign In or Register to comment.