Run selected test - #32
Conversation
|
Hello, I see hassfest failure, sorry for that. I can fixe it later on tomorrow. But I think that the fix for the malformed translation file is/was in my other PR. Please let me know. Thank you. |
|
No worries I've fixed already and released it |
|
Thank you very much and I am sorry again. There seems to be still minor (not affecting users) bug, where even if only card for the given speed test instance trigger the test, all speed test are triggered one by one. I wiloild like to look into it today or tomorrow. |
|
Sure no worries, I'm sorry I don't have much time to put into it these days, I've just moved to a new country and got realy busy and working on other integrations too. |
|
Hey no problem at all, you are super fast with the merge, no issues at all. Thank you for this project. I tested other integrations and this one is the only one working fine and covering all my needs. Kudos! |
|
I want to query something with you, some users reported that download speed is not correct, is yours fine? |
|
I have little bit special installation - I still run supervised on the beefy server, so I am not limited by the hardware. I have semi professional network setup based on Mikrotik infrastructure and using 5009 main router. I have fiber 100/100, VDSL 50/10 and LTE 25/5 connections. All WANs are measuring as expected when compared to the locally runner ookla test and also corresponding to the ISP declared values on both upload and download speeds (within +5-10pct). So I can't confirm such reports as my speeds are always overall better than declared speeds, both downloads and uploads. I can rerun few test later today but I do not expect any difference. I just quickly run all test from HA and it is as I said, no issues. |
|
Alright thanks for the info |
Currently the run_speedtest service trigger run of all configured test at the same time. This is not desired, as user might want to select only given test or execute test in series not in parallel as it might affect speed test results.
This change implements optional selector of the device_id o be specified for run_speedtest. If not specified test will be executed not in parallel but in serial manner.
This PR was previously implemented on the top of my PR #31, there might be some conflicts. If would would merge #31 I can recreate this PR against it fro clean merge.