BigIP Report Old

Hi Mihir

Sorry for the late reply. I did not get a notification when you made this comment. Will poke the DC team about that.

To answer your question. No I have never had the reason to do that as most places I've worked at a longer time did not have GTM.

/Patrik

New version out now. Calling it a Beta for a while since the code has only been tested in two places by two people (Tim Riker and myself). BIG thanks to Tim for contributing the code!

Since I did not do that much in this release other than reviewing and testing I'll let Tim describe the new features and fixes!

Take it away  !

Version 5.3.0 is now available.

http://loadbalancing.se/downloads/bigipreport-5.3.0-beta.zip

This release includes bug fixes and a number of new features. You will need to update your xml config files to include the new settings.

• Regular Expression searching on both general and column searches. Some examples in the help
• rename "underlay" instead of "Move the content of this folder to the configured report root"
• new settings in xml file! See example and add autoExpandPools, showAdcLinks, regexSearch
• new preferences.json containing preferences including polling settings
• datagroup caching now uses the REST service. the caching account on the F5 for versions older than 12 will need to be an Administrator. For APM retrieval the account will need to be an Administrator even on newer versions of BIG-IP.
• avoid status polling for the same pool multiple times each poll cycle
• stats to loggederrors, more entries show up on the Log tab.
• hide some columns by default on various tables
• links in datagroups to pools and http/https entries
• snat pool fixes, allow filtering by snat pool
• new status searching, see help. {ENABLED|DISABLED}:{BLUE|GREEN|RED}
• updated tab/button/input styling. Now includes row hover coloring, and sort column coloring. Note: Datatables supports sorting on multiple columns.
• expand/collapse support for associated pools, datagroups, virtual servers, etc. Not supported on Edge.
• expand/collapse/restore button to change current view
• monitor column on pool table, allows searching pools by monitor name
• column search boxes have clear (x) on Chrome. Not supported on Edge or Firefox.

Great tool. Sorry if this was already asked, but is there any plan to add association between LTM policies and pools? Kinda like the replacement for iRule association to pools.

In the upcoming versions we'll migrate from the SOAP api and onto the REST API. This will enable us to integrate policies. Stay tuned. :)

Great work on this! I really like how it is able to interpret pools from iRules. That is something I struggled with in my own attempts at something similar.

One issue I am having is when I configure a relatively large amount LTMs in the XML file the web page has issues loading. I have around 60 or so devices and the json files that are created after running the script are large. The webpage will usually load up to 99% and then sit for awhile. If you wait a few minutes it eventually loads. However it seems to only be able to load part of the page/content. The HTML looks normal. JSON files also look correct. It may be an issue with my NGINX settings though so I am starting there. If I test with a single LTM it works fine.

  What are your stats? Originally BigIPReport rendered all content into the html file, and then loaded all the json files to handle searching and linking etc. Now we only render content for each tab in the client when you switch to the tab. This has make the loading much faster. We do wait until all the json files are loaded before rendering the initial screen to insure that linking and cross references all work. How large are your files? In my environment on 5.3.0 (plus some fixes) I have this:

Stats: G:21 LB:34 VS:5961 P:6486 R:5004 DG:756 C:483 M:4170 ASM:219 T:1605.7656189

Search for "stats" on the "log" tab to see this output. You can search for "loaded:" to get this:

loaded: loadbalancers:34, virtualservers:5961, pools:6486, iRules:5004, datagroups:756, certificates:483, monitors:4170, asmPolicies:219.

which is rendered on the client once the json files have loaded. Looking at the previous message, I see it took 6 seconds to load and render for me. I'm on a local fast connection. It takes much longer to load and render while remote over vpn. My pools.json is around 14M in size. That's the largest one.

Side note: we also track pools when linked from datagroups. ie: virtual-server -> irule -> datagroup -> pool We do not currently track datagroup names or pool names that are assembled in an irule. ie: dg_$vip\_something so the text names of datagroups or pools need to actually occur in the irule. They can be in a comment, and will still be linked. Sometimes I consider that a bug, and sometimes a feature.

 Thanks for the quick response. I am running Version 5.2.8

I do not see stats in my logs. I removed a bunch of devices to test but I am seeing the same issue with only 6 LTMs.

loaded: loadbalancers:6, virtualservers:1732, pools:2426, iRules:397, certificates:1472, datagroups:68, asmPolicies:0.

If I setup an IIS server on my local box it loads very quickly (under 5 seconds). If I export all files to a remote Nginx webserver it either fails to load or takes several minutes. This is why I am assuming there is some issue with my Nginx settings. In the browser I can see it taking several minutes to try to load the specific pools and virtual json files.

 can you download the files from your nginx manually? Something like:

$ curl https://loadbalancing.se/bigipreportdemo/json/pools.json

I don't think we rely on the mime type, but I'm not positive on that. The images, css, javascript, and json files need to all be readable. If your small sample works, I would think the permissions are correct. You're running the small sample from the ngnix server too?

Note: My current 5.3.1 snapshot runs over REST instead of SOAP and does not need the pssnapin. I needs powershell 6.x or newer, and can run directly on non-windows machines.

 correct I can download all the json files manually and navigate to them as well. The permissions seem to be correct. And yes I have tested with a single small dev F5 inside nginx and it works fine. I'll continue to tweak nginx as I really think that is the bottleneck.