I consistently see that malformed URI error in both passing and failing tests so I think it's a red herring. That said, we do have an issue where the host is `"None"` from time to time. I noticed that the mozharness `--host` setting doesn't have a default setting so that might be the root cause of it but I haven't dug into that too much; :whimboo's patch will help us debug this further.
It seems like the error `GeckoConsole: [JavaScript Error: "IndexedDB UnknownErr: ActorsParent.cpp:613"]` is the real one here. I haven't been able to find out why it happens, but I managed to find that it's caused by clearing the extensions local storage so it's not related to what we are setting:
```
03-06 18:14:31.753 8014 8029 I Gecko : console.info: "[raptor-runnerjs] attempting to write to local storage"
03-06 18:14:31.754 8014 8029 I GeckoConsole: [raptor-runnerjs] attempting to write to local storage
03-06 18:14:31.754 8014 8029 I Gecko : console.info: "[raptor-runnerjs] [object Object]"
03-06 18:14:31.755 8014 8029 I GeckoConsole: [raptor-runnerjs] [object Object]
03-06 18:14:31.788 8014 8080 E GeckoConsole: [JavaScript Error: "IndexedDB UnknownErr: ActorsParent.cpp:613"]
03-06 18:14:31.789 8014 8029 E GeckoConsole: [JavaScript Error: "UnknownError: The operation failed for reasons unrelated to the database itself and not covered by any other error code." {file: "resource://gre/modules/ExtensionStorageIDB.jsm" line: 812}]
```
This was found using this debugging patch - had we cleared the extension data we would have seen a "cleared storage" log message: https://treeherder.mozilla.org/#/jobs?repo=try&revision=bcc0fce02e7fdda70b032ed30903018e318bc76f&selectedJob=292022290
I don't see a reason why we might be clearing the extension storage since the `set()` function overwrites existing values with the new ones. Maybe this clearing is needed to get rid of settings which are not in the new settings. In either case, I'm testing to see what happens if we remove this clear here - it works locally without any problems: https://treeherder.mozilla.org/#/jobs?repo=try&revision=089bd80b6ca40d2c8434c1a9d94a4b539993ee21
Bug 1609295 Comment 95 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
I consistently see that malformed URI error in both passing and failing tests so I think it's a red herring. That said, we do have an issue where the host is `"None"` from time to time. I noticed that the mozharness `--host` setting doesn't have a default setting so that might be the root cause of it but I haven't dug into that too much; :whimboo's patch will help us debug this further.
It seems like the error `GeckoConsole: [JavaScript Error: "IndexedDB UnknownErr: ActorsParent.cpp:613"]` is the real one here. I haven't been able to find out why it happens, but I managed to find that it's caused by clearing the extensions local storage so it's not related to what we are setting:
```
03-06 18:14:31.753 8014 8029 I Gecko : console.info: "[raptor-runnerjs] attempting to write to local storage"
03-06 18:14:31.754 8014 8029 I GeckoConsole: [raptor-runnerjs] attempting to write to local storage
03-06 18:14:31.754 8014 8029 I Gecko : console.info: "[raptor-runnerjs] [object Object]"
03-06 18:14:31.755 8014 8029 I GeckoConsole: [raptor-runnerjs] [object Object]
03-06 18:14:31.788 8014 8080 E GeckoConsole: [JavaScript Error: "IndexedDB UnknownErr: ActorsParent.cpp:613"]
03-06 18:14:31.789 8014 8029 E GeckoConsole: [JavaScript Error: "UnknownError: The operation failed for reasons unrelated to the database itself and not covered by any other error code." {file: "resource://gre/modules/ExtensionStorageIDB.jsm" line: 812}]
```
This was found using this debugging patch - had we cleared the extension data we would have seen a "cleared storage" log message: https://treeherder.mozilla.org/#/jobs?repo=try&revision=bcc0fce02e7fdda70b032ed30903018e318bc76f&selectedJob=292022290
I've read that some people have seen this `UnknownErr` error when the disk drives are faulty but that's not happening here since all of our devices have this error and it seems to be spread quite uniformly across all of them. The failure does occur much more often on G5 than P2 so it's possible that something specific to the devices is making the error more prominent.
I don't see a reason why we might be clearing the extension storage since the `set()` function overwrites existing values with the new ones. Maybe this clearing is needed to get rid of settings which are not in the new settings. In either case, I'm testing to see what happens if we remove this clear here - it works locally without any problems: https://treeherder.mozilla.org/#/jobs?repo=try&revision=089bd80b6ca40d2c8434c1a9d94a4b539993ee21