Skip to main content

Server error 500 and packet loss

InteractM
Is there any dependence between packet loss and server error 500? I have checked Apache access logs for 500s and it looks like random occurrence without any major cause. Also I have noticed that outer switch was generating collisions which may have correlation to packet loss on that network. In most cases it looks like somebody got 500 and then reload page and page got loaded. Also all 500 errors aren't equal - how can I be more specific by analyzing access log? Thanks

Comments

14 comments

Date Votes
  • InteractM
    What do I need to enable to stack trace those 500's errors?
    0
  • 24x7server
    Hello, What error you getting in your Apache error logs file ? Does you own a vps or a dedicated server ?
    0
  • InteractM
    It is a dedicated server An example Apache access log with 500 error: 157.55.39.112 - - [25/Sep/2015:10:20:45 -0400] "GET /meet-the-staff-finance-administration-technology-ud-3173/ HTTP/1.1" 500 70432 "-" "Mozilla/5.0 (compatible; bingbot/2.0; +
    0
  • Jcats
    You could try this:
    0
  • InteractM
    You could try this:
    0
  • Jcats
    Weird, you don't have any mod_sec or .htaccess rules used to block bots maybe? I know it would in most cases throw 403's but I have seen 500/501 in some cases.
    0
  • InteractM
    Not for my knowledge but I can compare other site on same server if has same problems with crawlers
    0
  • InteractM
    Well - other websites on the same sever aren't experience that problem
    0
  • Jcats
    Well if its a .htaccess rule than it wouldn't affect the others sites so it makes sense this could be the problem.
    0
  • InteractM
    .htaccess has a lot of rewrite rules for redirects (about 2.5k) as well some modifications from WP plugins.
    0
  • Jcats
    Well depending on how often the bots are hitting the site and throwing a 500 you can rename the .htaccess to something else so it stops triggering then wait for the next bot to stop by to see if it throws a 500. Otherwise, you can log into Google webmaster and you should be able to force Google to crawl the site which you can use to test it quicker so the site isn't sitting broken if it heavily relies on the .htaccess
    0
  • cPanelMichael
    Hello :) Have you been able to confirm if the "500" error messages are still occurring since increasing the output_buffering value in the php.ini file? Thank you.
    0
  • InteractM
    Yes - I still have those 500's even increasing output_buffering. Also I'm not able to stack trace most of those requests with 500's neither to PHP or MySQL errors logs as well Apache error logs. Example of 500's from Apache access log:
    41.79.186.247--[06/Oct/2015:11:23:49-0400]"GET /international-chamber-of-commerce-icc-ud-754/ HTTP/1.1"500-"http://www.bing.com""Mozilla/5.0 (Windows NT 6.2; rv:42.0) Gecko/20100101 Firefox/42.0" 54.193.63.98--[06/Oct/2015:11:29:05-0400]"GET / HTTP/1.1"500-"-""webmon-webagent v1.3" 199.167.17.37--[06/Oct/2015:11:39:23-0400]"GET / HTTP/1.1"500-"-""Mozilla/5.0 (Anturis Agent)"
    and appear that is relate only to one hosting account (one site) which runs Wordpress Another thing what I have done it was execute that code in the console:
    export USE_ZEND_ALLOC=0
    what have reduced 500's from around 100 per day to something below 20. Also I have noticed that those 500's are showing up after Apache auto restart (no clue why).
    0
  • cPanelMichael
    You could try switching to a different PHP handler or a different MPM to see if that makes a difference: MPM Options - EasyApache - cPanel Documentation Thank you.
    0

Please sign in to leave a comment.

Didn't find what you were looking for?

New post