I have set up an AGW to send traffic to an ACA using the same working configuration as in this issue: #65
I'm confused as to why I get 502 gateway errors if I remove the custom probe and instead use the default probe.
My understanding is that the default probe will make a request to the URL in the rule's backend pool. That seems intuitive to me.
And I expect that I could just use the default probe instead since the url in the custom probe is the same as in the backend pool.
But in the Azure documentation on troubleshooting AGW 502s, the section about Problems with default health probe, it says the default url for the default probe is 127.0.0.1. I'm not sure if the probe is really making a request using the loopback address, I imagine that would never work, or if it is just a placeholder for the backend pool's URLs. If it really is making a request for 127.0.0.1, then I get why you'd need a custom health probe to change that address.
I'd like to avoid having to specify a custom health probe. Eventually for my 5 ACAs I'll have 5 listeners, 5 backend pools, etc.. That's okay, but I don't want to have to also create 5 custom health probes and 5 backend settings. I noticed each custom probe can only be associated with one backend setting, which means I'll need a different backend setting for each of the 5 custom probes so they can probe different URLs.
Sorry if this is the wrong place to ask this question. While troubleshooting this issue I found #65 and figured that asking a question clarifying the kind of info shared in that thread would be appropriate.
I have set up an AGW to send traffic to an ACA using the same working configuration as in this issue: #65
I'm confused as to why I get 502 gateway errors if I remove the custom probe and instead use the default probe.
My understanding is that the default probe will make a request to the URL in the rule's backend pool. That seems intuitive to me.
And I expect that I could just use the default probe instead since the url in the custom probe is the same as in the backend pool.
But in the Azure documentation on troubleshooting AGW 502s, the section about Problems with default health probe, it says the default url for the default probe is
127.0.0.1. I'm not sure if the probe is really making a request using the loopback address, I imagine that would never work, or if it is just a placeholder for the backend pool's URLs. If it really is making a request for127.0.0.1, then I get why you'd need a custom health probe to change that address.I'd like to avoid having to specify a custom health probe. Eventually for my 5 ACAs I'll have 5 listeners, 5 backend pools, etc.. That's okay, but I don't want to have to also create 5 custom health probes and 5 backend settings. I noticed each custom probe can only be associated with one backend setting, which means I'll need a different backend setting for each of the 5 custom probes so they can probe different URLs.
Sorry if this is the wrong place to ask this question. While troubleshooting this issue I found #65 and figured that asking a question clarifying the kind of info shared in that thread would be appropriate.