Getting total bandwidth currently utilized in an iRule

Question

Hope someone can help.&nbsp;
 &nbsp;
I have finite internet bandwidth available. There are times when there are too many incoming user requests. I would like to turn away new users (new login requests) based on the amount of internet bandwidth that is already utilized.&nbsp;
 &nbsp;
I can think of the following ways to achieve this - but they all get stuck at one point - my iRule needs to know the total bandwidth being utilized so that it can flash the "server busy" page to new users.&nbsp;
 &nbsp;
Method 1&nbsp;
A simple virtual server with the iRule attached. The iRule can compare the current bandwidth (incoming or outgoing) being utilized by this virtual server. If it is greater than my a pre-defined threshold, the iRule will respond with a "Server Busy" message. This way, new users will not get into the system while users already logged-in will continue to get good experience.&nbsp;
 &nbsp;
Method 2&nbsp;
Apply a Rate Class to the virtual server. This will ensure the pre-defined bandwidth limit is not exceeded. But unless the above mentioned iRule is applied, new users as well as already-logged-in users will see deteriorating performance.&nbsp;
To make the iRule work, I would need to know the cumulative bandwidth being consumed by all connections that have this Rate Class applied.&nbsp;
 &nbsp;
Method 3&nbsp;
Very similar to Method 2 - but using Bandwidth Controller instead. Same problem - need to know the cumulative bandwidth being utilized by all connections that have this Bandwidth Controller applied.&nbsp;
I was hoping the BWC class would have the required information available directly.&nbsp;
 &nbsp;
thanks,&nbsp;
MM&nbsp;

nitass · Answer

not sure if this is useful. 
  
 since 11.1.0, virtual server or pool AVR statistics can be obtained using ISTATS::get command in irule (e.g. ISTATS::get "VIP virtual-server-name counter type). anyway, it is not realtime statistics because statistics needs at least about 10 seconds to be updated (i.e. it is previous update interval statistics). to see all the types, you can run "istats dump" in bash. 
  
 e.g. 
 root@(B3900-R67-S12)(cfg-sync Standalone)(Active)(/Common)(tmos) list ltm virtual bar
ltm virtual bar {
    destination 100.100.100.101:80
    ip-protocol tcp
    mask 255.255.255.255
    pool foo
    profiles {
        http { }
        myanalytics { }
        tcp { }
    }
    rules {
        myrule
    }
    source 0.0.0.0/0
    source-address-translation {
        type automap
    }
    vs-index 3
}
root@(B3900-R67-S12)(cfg-sync Standalone)(Active)(/Common)(tmos) list ltm profile analytics myanalytics
ltm profile analytics myanalytics {
    app-service none
    collect-geo enabled
    collect-ip enabled
    collect-max-tps-and-throughput enabled
    collect-methods enabled
    collect-page-load-time enabled
    collect-response-codes enabled
    collect-url enabled
    collect-user-agent enabled
    collect-user-sessions disabled
    defaults-from analytics
    description none
    notification-email-addresses none
    partition Common
    session-cookie-security ssl-only
    session-timeout-minutes 5
    traffic-capture {
        capturing-for-myanalytics {
            captured-protocols all
        }
    }
}
root@(B3900-R67-S12)(cfg-sync Standalone)(Active)(/Common)(tmos) list ltm rule myrule
ltm rule myrule {
    when HTTP_REQUEST {
  if { [HTTP::uri] eq "/stats" } {
    log local0. "request_throughput: [ISTATS::get "VIP [virtual name] counter request_throughput"]"
    log local0. "response_throughput: [ISTATS::get "VIP [virtual name] counter response_throughput"]"
  }
}
}

test

[root@B3900-R67-S12:Active:Standalone] config  tail -f /var/log/ltm
Aug  1 09:37:47 B3900-R67-S12 info tmm3[9398]: Rule /Common/myrule : request_throughput: 7680
Aug  1 09:37:47 B3900-R67-S12 info tmm3[9398]: Rule /Common/myrule : response_throughput: 339520

istats dump

[root@B3900-R67-S12:Active:Standalone] config  istats dump
/var/tmstat/cluster.tm2 [READONLY]: nextBlockId=4064, last update at 2013-08-01 09:39:08

all facts:
   [ GEO=N/A ][concurrent_sessions] = 0 (2013-08-01 09:39:08)
   [ GEO=N/A ][new_sessions] = 0 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][page_load_time_samples] = 0 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][request_throughput] = 5760 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][response_throughput] = 254640 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][page_load_time] = 0 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][concurrent_sessions] = 0 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][max_page_load_time] = 0 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][server_latency] = 68 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][tps] = 41 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][max_server_latency] = 3000 (2013-08-01 09:39:08)
   [ VIP=/Common/bar ][new_sessions] = 0 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][page_load_time_samples] = 0 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][request_throughput] = 5760 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][response_throughput] = 254640 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][page_load_time] = 0 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][max_page_load_time] = 0 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][server_latency] = 68 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][tps] = 41 (2013-08-01 09:39:08)
   [ POOL=200.200.200.101:80 ][max_server_latency] = 3000 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][page_load_time_samples] = 0 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][request_throughput] = 0 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][response_throughput] = 0 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][page_load_time] = 0 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][max_page_load_time] = 0 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][server_latency] = 0 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][tps] = 0 (2013-08-01 09:39:08)
   [ vip=/Common/bar ][max_server_latency] = 70 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][page_load_time_samples] = 0 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][request_throughput] = 0 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][response_throughput] = 0 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][page_load_time] = 0 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][max_page_load_time] = 0 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][server_latency] = 0 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][tps] = 0 (2013-08-01 09:39:08)
   [ pool=200.200.200.101:80 ][max_server_latency] = 70 (2013-08-01 09:39:08)

maheshme973_569 · Answer

Thanks nitass! Looks like something I can use. I'll try it out.

maheshme973_569 · Answer

Thanks nitass! Looks like something I can use. I'll try it out. And 10 seconds is good enough for me.

maheshme973_569 · Answer

Finally found time to try it out. Looks like the AVR module is required. I don't have it. Bummer! 
&nbsp;  
&nbsp; Any other way to get the throughput information for a virtual server?

maheshme973_569 · Answer

Wish I could recall the previous post. I do have the AVR module - it was not provisioned. I will be doing it tonight.&nbsp;
 &nbsp;
On a related note, we are running 11.3.0 - so enabling throughput and tps collection is slightly different. It has to be done from tmsh.&nbsp;
http://support.f5.com/kb/en-us/prod...mlnewfeat&nbsp;

Forum Discussion

Getting total bandwidth currently utilized in an iRule

6 Replies

Recent Discussions

snmp automatic stop mail send

BIG-IP DNS iRule issue with static variable

Using okta as SSO to login F5 GUI

(How) can I get two client certificates in one APM session?

BIG-IP DNS: Check Status Of Multiple Monitors Against Pool Member

Related Content

How much bandwidth are we using?

ISP Bandwidth Utilization

Exceed bandwidth utilization on standby device

F5 utilization

How Much Bandwidth is Too Much?