NSX Load Balancer - Test Plan

[rajeevsrikant]

Have any one has done the evaluation of the NSX load balancer with some proper test plan.

If so can you please share the same like what are the functional and non functional tests which have been done.

Basically i need a test plan to test the features and record it.

[rajeevsrikant]

Any body who has done NSX load balancer testing could you please share your inputs ?

[remo_williams]

Deploy One-Arm Proxy

1. Deploy ESG as standalone on same tier where your apps will live (web/app whatever)

2. Create self-signed cert, or import existing cert.

3. Create your APP Profiles, health monitors, Pools

4. Create a VIP using a secondary IP address.

Test

1. Best tested with 2 different client endpoints (e.g. separate physical or VM)

2. Connect to URL from both machines.

3. From command line of Load Balancer run the following command.

     "show service loadbalancer pool"

4. Pay attention to the output.

POOL https_pool

|  LB METHOD round-robin

|  LB PROTOCOL L7

|  Transparent disabled

|  SESSION (cur, max, total) = (0, 1, 7)

|  BYTES in = (2453), out = (295312)

   +->POOL MEMBER: https_pool/tenant-01-https-1, STATUS: UP

   |  |  HEALTH MONITOR = BUILT-IN, default_http_monitor:L7OK

   |  |  |  LAST STATE CHANGE: 2017-01-19 21:02:48

   |  |  SESSION (cur, max, total) = (0, 1, 4)

   |  |  BYTES in = (1404), out = (195924)

   +->POOL MEMBER: https_pool/tenant-01-https-2, STATUS: UP

   |  |  HEALTH MONITOR = BUILT-IN, default_http_monitor:L7OK

   |  |  |  LAST STATE CHANGE: 2017-01-19 21:02:48

   |  |  SESSION (cur, max, total) = (0, 1, 3)

   |  |  BYTES in = (1049), out = (99388)

If you are seeing 0's for BYTES in / out, for any pool member then there's potentially an issue. This data will change, so as you refresh the session re-run command to see other values. The other show commands are helpful too, but this command giving good results is sufficient in my opinion.

[rajeevsrikant]

Thanks

Is there any alerting or notification which can be generated from the NSX LB in case if any one of the pool member is down or when the http service of any of the pool member is down ?

I understand this can be achieved by doing server side monitoring. But is there any function in NSX LB in which it detects it and sends notifications.

[ddesmidt]

When backend pool member is down and/or the whole VIP is down; the Edge will generate a syslog message.

Dimitri

[rajeevsrikant]

Thanks

But currently there is no e-mail notifications or alerting right directly from NSX

[cnrz]

Email notifications can be configured for NSX Load Balancer (and other Edge Services) through Loginsight (which is no cost for NSX).

http://www.virten.net/2016/05/integrate-vmware-nsx-in-vmware-log-insight/

https://pubs.vmware.com/log-insight-25/index.jsp?topic=%2Fcom.vmware.log-insight.administration.doc%...

Administrators can configure vRealize Log Insight to send email notifications when certain events occur in the system. The from address of system notification emails is configured by the administrator user on the SMTP configuration page of the Administration UI, in the Sender text box. See Configure the SMTP Server for vRealize Log Insight.

Administrator users can also configure Log Insight to send notification emails when the storage capacity drops below a defined threshold.

Every vRealize Log Insight user can create alert queries to receive email notifications from vRealize Log Insight when certain criteria are met.

Dashboards for monitoring Pool members status is also possible with VROPS.

Sample dashboards to monitor NSX Load Balancershttp://virtual-red-dot.info/sample-dashboards-to-moni...

[rajeevsrikant]

Thanks

In addition to this is there any way I can do performance test in terms resource utilization, no of connections / sessions.

I don't have any separate tool to do performance test. But how can i do it with out that.