Crash in [@ mozilla.appservices.fxaclient.FxaException$Network: at mozilla.appservices.fxaclient.rust.RustError.intoException(RustError.kt:5)]

[mstange]

In the last day, the rate of crashes with this signature has gone up considerably. Maybe a regression from #7856? @grigoryk

It sounds similar to #7536 which was about not handling network failures, if I understand things correctly.

Crash report: https://crash-stats.mozilla.org/report/index/92c220aa-3be3-4c80-b071-af3e30200924

Java stack trace:

mozilla.appservices.fxaclient.FxaException$Network
	at mozilla.appservices.fxaclient.rust.RustError.intoException(RustError.kt:5)
	at mozilla.appservices.fxaclient.FirefoxAccount.getTokenServerEndpointURL(FirefoxAccount.kt:9)
	at mozilla.components.service.fxa.FirefoxAccount.getTokenServerEndpointURL(FirefoxAccount.kt:1)
	at mozilla.components.service.fxa.manager.FxaAccountManager.maybeUpdateSyncAuthInfoCache(FxaAccountManager.kt:13)
	at mozilla.components.service.fxa.manager.FxaAccountManager$maybeUpdateSyncAuthInfoCache$1.invokeSuspend(Unknown Source:11)
	at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:3)
	at kotlinx.coroutines.DispatchedTask.run(DispatchedTask.kt:21)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1167)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:641)
	at java.lang.Thread.run(Thread.java:919)

┆Issue is synchronized with this Jira Task

[grigoryk]

Given timing, most likely a regression from #7856. Here's a Sentry report with a bit more detail, showing the same spike: https://sentry.prod.mozaws.net/operations/firefox-nightly/issues/8905012/?query=is%3Aunresolved

mozilla.appservices.fxaclient.Network: Network error: [no-sentry] Network error: Java error: "fetch error: org.mozilla.geckoview.WebRequestError: Request failed, error=0x25, category=0x5"
    at mozilla.appservices.fxaclient.rust.RustError.intoException(RustError.kt:5)
    at mozilla.appservices.fxaclient.FirefoxAccount.getTokenServerEndpointURL(FirefoxAccount.kt:9)
    at mozilla.components.service.fxa.FirefoxAccount.getTokenServerEndpointURL(FirefoxAccount.kt:1)
    at mozilla.components.service.fxa.manager.FxaAccountManager.maybeUpdateSyncAuthInfoCache(FxaAccountManager.kt:13)
    at mozilla.components.service.fxa.manager.FxaAccountManager$maybeUpdateSyncAuthInfoCache$1.invokeSuspend
    at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:3)
    at kotlinx.coroutines.DispatchedTask.run(DispatchedTask.kt:21)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1167)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:641)
    at java.lang.Thread.run(Thread.java:919)

[grigoryk]

We're not handling network failures within getTokenServerEndpointURL, which could happen if remote config needs to be populated within the fxaclient (on startup?).

It's not obvious to me why this is spiking now, as far as I can tell prior to #7856 we were just as vulnerable to this, and I don't think behaviour of getTokenServerEndpointURL changed recently within appservices either.

Either way, patch incoming to handle these failures gracefully.

[mstange]

Looks like it worked! The graph went back down.

[grigoryk]

Excellent!