#106 bug report

Closed
opened 1 year ago by bugreporter999 · 1 comments

Opened my fire tablet today and saw the notification asking me to share a bug report, so here it is. WiFi was off at this time but I have never had an error before With FFUpdater from leaving a device disconnected from the internet.

Stacktrace:

de.marmaro.krt.ffupdater.app.impl.exceptions.NetworkException: Fail to consume 'https://firefox-ci-tc.services.mozilla.com/api/index/v1/task/mobile.v2.fenix.release.latest.armeabi-v7a/artifacts/public/logs/chain_of_trust.log'.
    at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer$consumeAsync$2$1.invokeSuspend(Unknown Source:35)
    at r3.a.resumeWith(Unknown Source:8)
    at f4.e0.run(Unknown Source:86)
    at i4.a.p(Unknown Source:0)
    at i4.a$a.run(:5)
Caused by: java.net.UnknownHostException: Unable to resolve host "firefox-ci-tc.services.mozilla.com": No address associated with hostname
    at java.net.Inet6AddressImpl.lookupHostByName(Inet6AddressImpl.java:157)
    at java.net.Inet6AddressImpl.lookupAllHostAddr(Inet6AddressImpl.java:105)
    at java.net.InetAddress.getAllByName(InetAddress.java:1154)
    at t2.e.b(Unknown Source:5)
    at o4.d.a(:30)
    at o4.a.intercept(:11)
    at p4.f.b(:3)
    at m4.a.intercept(:22)
    at p4.f.b(:3)
    at p4.a.intercept(:16)
    at p4.f.b(:3)
    at p4.h.intercept(:25)
    at p4.f.b(:3)
    at o4.e.f(:10)
    at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer.consume(:9)
    at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer.access$consume(Unknown Source:0)
    at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer$consumeAsync$2$1.invokeSuspend(Unknown Source:13)
    ... 4 more
Caused by: android.system.GaiException: android_getaddrinfo failed: EAI_NODATA (No address associated with hostname)
    at libcore.io.Linux.android_getaddrinfo(Native Method)
    at libcore.io.BlockGuardOs.android_getaddrinfo(BlockGuardOs.java:172)
    at java.net.Inet6AddressImpl.lookupHostByName(Inet6AddressImpl.java:137)
    ... 20 more

Device information: | key | value | | --- | --- | | version name | 75.5.0 | | version code | 102 | | build type | release | | SDK version | 28 | | android release | 9 | | board | onyx | | brand | Amazon | | device | onyx | | manufacturer | Amazon | | model | KFONWI | | product | onyx | | SOC manufacturer | / | | SOC model | / | | supported ABIs | armeabi-v7a, armeabi | | tags | amz-p,release-keys | | build time | 1650818947000 |

Opened my fire tablet today and saw the notification asking me to share a bug report, so here it is. WiFi was off at this time but I have never had an error before With FFUpdater from leaving a device disconnected from the internet. Stacktrace: ``` de.marmaro.krt.ffupdater.app.impl.exceptions.NetworkException: Fail to consume 'https://firefox-ci-tc.services.mozilla.com/api/index/v1/task/mobile.v2.fenix.release.latest.armeabi-v7a/artifacts/public/logs/chain_of_trust.log'. at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer$consumeAsync$2$1.invokeSuspend(Unknown Source:35) at r3.a.resumeWith(Unknown Source:8) at f4.e0.run(Unknown Source:86) at i4.a.p(Unknown Source:0) at i4.a$a.run(:5) Caused by: java.net.UnknownHostException: Unable to resolve host "firefox-ci-tc.services.mozilla.com": No address associated with hostname at java.net.Inet6AddressImpl.lookupHostByName(Inet6AddressImpl.java:157) at java.net.Inet6AddressImpl.lookupAllHostAddr(Inet6AddressImpl.java:105) at java.net.InetAddress.getAllByName(InetAddress.java:1154) at t2.e.b(Unknown Source:5) at o4.d.a(:30) at o4.a.intercept(:11) at p4.f.b(:3) at m4.a.intercept(:22) at p4.f.b(:3) at p4.a.intercept(:16) at p4.f.b(:3) at p4.h.intercept(:25) at p4.f.b(:3) at o4.e.f(:10) at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer.consume(:9) at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer.access$consume(Unknown Source:0) at de.marmaro.krt.ffupdater.app.impl.fetch.ApiConsumer$consumeAsync$2$1.invokeSuspend(Unknown Source:13) ... 4 more Caused by: android.system.GaiException: android_getaddrinfo failed: EAI_NODATA (No address associated with hostname) at libcore.io.Linux.android_getaddrinfo(Native Method) at libcore.io.BlockGuardOs.android_getaddrinfo(BlockGuardOs.java:172) at java.net.Inet6AddressImpl.lookupHostByName(Inet6AddressImpl.java:137) ... 20 more ``` Device information: | key | value | | --- | --- | | version name | 75.5.0 | | version code | 102 | | build type | release | | SDK version | 28 | | android release | 9 | | board | onyx | | brand | Amazon | | device | onyx | | manufacturer | Amazon | | model | KFONWI | | product | onyx | | SOC manufacturer | / | | SOC model | / | | supported ABIs | armeabi-v7a, armeabi | | tags | amz-p,release-keys | | build time | 1650818947000 |
Tobiwan2 commented 1 year ago

(notabug.org is currently not accepting my 2nd login factor - that's why I'm using this account)

Version 77.0.3 should improve this situation.

My goal is for users to be able to rely on the background check. Therefore, I try not to hide errors that occur in the background. But smartphones loose their internet connection often - so it is not a solution to always display every error.

My previous solution was: If an network error occurs, then retry up to 8/19 times. If the 8/19 retry was unsuccessful, then the last error will be displayed as a notification.

(8 retries -> total of 1 hour -> wait 15s after first try, 30s after second, 1m after third, 2m, 4m, 8m, 16m, 32m) (19 retries -> total of 2 days -> wait 15s after first try, 30s after second, 1m after third, 2m, 4m, 8m, 16m, 32m, 64m, 128m, 256m, 300m, 300m, 300m, 300m, 300m, 300m, 300m, 300m)

It is very difficult to detect if the user has internet enabled or disabled - that's why I don't try to detect it. For examples these and more combinations are possible: enabled airplane mode, enabled airplane mode + enabled wifi, disabled airplane mode + disabled wifi, enabled airplane mode + ethernet/lan.

In 77.0.3 I increase the max retries to 34 (5 days). If the last retry fails, FFUpdater will display a different notification with improved explanation about the kind of this "error".

(notabug.org is currently not accepting my 2nd login factor - that's why I'm using this account) Version 77.0.3 should improve this situation. My goal is for users to be able to rely on the background check. Therefore, I try not to hide errors that occur in the background. But smartphones loose their internet connection often - so it is not a solution to always display every error. My previous solution was: If an network error occurs, then retry up to 8/19 times. If the 8/19 retry was unsuccessful, then the last error will be displayed as a notification. (8 retries -> total of 1 hour -> wait 15s after first try, 30s after second, 1m after third, 2m, 4m, 8m, 16m, 32m) (19 retries -> total of 2 days -> wait 15s after first try, 30s after second, 1m after third, 2m, 4m, 8m, 16m, 32m, 64m, 128m, 256m, 300m, 300m, 300m, 300m, 300m, 300m, 300m, 300m) It is very difficult to detect if the user has internet enabled or disabled - that's why I don't try to detect it. For examples these and more combinations are possible: enabled airplane mode, enabled airplane mode + enabled wifi, disabled airplane mode + disabled wifi, enabled airplane mode + ethernet/lan. In 77.0.3 I increase the max retries to 34 (5 days). If the last retry fails, FFUpdater will display a different notification with improved explanation about the kind of this "error".
Sign in to join this conversation.
Loading...
Cancel
Save
There is no content yet.