Android.system.GaiException

[Jim Morris]

Hi all,
I've had a simple forward SMS to Slack automation running for months now, and recently I've seen this error start to pop up, which then requires me to manually restart the flow at the most inconvenient times. Wondering if someone can explain what this error is?
02-14 06:16:00.747 F 4@3: android.system.GaiException: android_getaddrinfo failed: EAI_NODATA (No address associated with hostname) 02-14 06:16:00.754 I 4@3: Stopped by failure

[Jim Morris]

Also, if anyone knows how to get the flow to auto-restart after this error has occurred, that would be really helpful too!

[мJ ZЖУ]

Use Fail Catch block I think this sometimes happen if connection drop somewhere you could also add delay (Delay Block) for like 10 minutes between it.

[Henrik "The Developer" Lindqvist]

Such failure is usually caused by some DNS error, maybe due to VPN usage, or just a lack of internet connectivity.

Use the Failure catch block to handle a failure.

[Jim Morris]

Yeah a delay isn't going to help, unfortunately, as certain messages that get received have a useable time limit on them. :(

Definitely no VPN being used either at home or work. 

Apologies in advance for all the dumb/noob questions; I'm very new to this stuff.

Ok, now I have the Fail Catch element, how do I use it?

Where do I stick it in the flow and how do I configure it?

This is the flow I have thus far:

This is the log since I last restarted it yesterday:

06-03 10:46:21.860 I 7@1: Flow beginning
06-03 10:46:21.863 I 7@2: SMS received
06-03 10:47:39.129 I 7@3: HTTP request
06-03 10:47:39.652 I 7@2: SMS received
06-03 16:24:50.712 I 7@3: HTTP request
06-03 16:24:51.183 I 7@2: SMS received
06-03 17:17:56.516 I 7@3: HTTP request
06-03 17:17:56.971 I 7@2: SMS received
06-04 06:58:21.989 I 7@3: HTTP request
06-04 06:58:22.007 F 7@3: android.system.GaiException: android_getaddrinfo failed: EAI_NODATA (No address associated with hostname)
06-04 06:58:22.023 I 7@3: Stopped by failure

[Henrik "The Developer" Lindqvist]

Place it before the block that's failing. To simply ignore the failure the flow can also use a Fork block:

1. Flow beginning

2. SMS received

3. Fork

3.1. HTTP request

3.2. (#3.1 should be unconnected)

4. (connect #3 OK to #2 IN) 

With or without an Failure catch block using a Fork block is preferable, since the HTTP request block may take several seconds therefor "blocking" the SMS received block.

[Jim Morris]

@henrik

Do you mean like this?


How does it know to send the HTTPrequest if it is not connected?

I have also tried this, which seems to be working atm, but I guess the real test will come if a failure event happens:
 

[мJ ZЖУ]

Try allow data in background for Automate from its app setting if you want to avoid sending HTTP request twice.

[Jim Morris]

I assume that this is the setting you are referring to, as I couldn't find one like it in the Automate app settings itself.

[мJ ZЖУ]

Yes that's should do it for meantime and try to save HTTP response to a file if no further error show then leave it at that.

[Henrik "The Developer" Lindqvist]

Connect the Fork NEW path to the HTTP request block.

In that configuration the simply ignores the failure for 5 retries, the fail.