Skip to content
This repository has been archived by the owner on Mar 28, 2024. It is now read-only.

[BUG-228624] [EEP] "Could not load the settings for [NAME] from the database" error #6658

Closed
sl-service-account opened this issue May 1, 2020 · 19 comments

Comments

@sl-service-account
Copy link

sl-service-account commented May 1, 2020

What just happened?

Just starting to create after EEP maintenance update.
So as the environment is too visually dark during the build, I would usually leave the used shared environment and select midday.
To my surprise lo and behold, presented with dialog "Could not load the settings for [NAME] from the database".

What were you doing when it happened?

Usually nothing as to what was stated above.

What were you expecting to happen instead?

I was not expecting the dialog.

Other information

No.

Attachments

Original Jira Fields
Field Value
Issue BUG-228624
Summary [EEP] "Could not load the settings for [NAME] from the database" error
Type Bug
Priority Unset
Status Closed
Resolution Unactionable
Labels whirly-eep
Reporter CeeEss (ceeess)
Created at 2020-05-01T03:36:20Z
Updated at 2020-06-15T15:23:00Z
{
  'Build Id': 'unset',
  'Business Unit': ['Platform'],
  'Date of First Response': '2020-05-01T12:28:02.543-0500',
  "Is there anything you'd like to add?": 'No.\r\n',
  'ReOpened Count': 0.0,
  'Severity': 'Unset',
  'System': 'SL Viewer',
  'Target Viewer Version': 'viewer-development',
  'What just happened?': 'Just starting to create after EEP maintenance update.\r\nSo as the environment is too visually dark during the build, I would usually leave the used shared environment and select midday.\r\nTo my surprise lo and behold, presented with dialog "Could not load the settings for [NAME] from the database".\r\n',
  'What were you doing when it happened?': 'Usually nothing as to what was stated above.',
  'What were you expecting to happen instead?': 'I was not expecting the dialog.\r\n',
  'Where': 'You are at 224.9, 162.9, 28.3 in Deneb located at sim10479.agni.lindenlab.com (216.82.51.185:13004)\r\n',
}
@sl-service-account
Copy link
Author

CeeEss commented at 2020-05-01T04:27:54Z

Tested it.  Works as expected on Sunrise, Sunset and Midnight but not on Midday.

 

@sl-service-account
Copy link
Author

CeeEss commented at 2020-05-01T04:41:43Z

Hmm.  After playing around a little bit with the settings the server finally awakened.

Midday environment is restored.

 

@sl-service-account
Copy link
Author

Dan Linden commented at 2020-05-01T17:28:03Z

Thank you, CeeEss.

If this happens again, please attach your SecondLife.log file. This page tells you where to find the logs folder: http://community.secondlife.com/t5/English-Knowledge-Base/How-to-report-a-bug/ta-p/733545#Section_.3

Dan

@sl-service-account
Copy link
Author

Whirly Fizzle commented at 2020-05-01T23:10:18Z, updated at 2020-05-01T23:18:39Z

This is definitely a bug on EEP viewers.
We've seen the same problem on internal Firestorm EEP viewers.
I'm afraid I don't know the repro steps though - I haven't been able to reproduce it myself yet, but several of our testers seem to be able to reproduce it regularly.
The bug can occur seemingly with any environment, not just default midday, howver it's random.
The "Could not load the settings for [NAME] from the database" error message is most usually seen seconds after login completes.

Error on FS EEP: https://prnt.sc/rwznq2
Error in log on FS EEP: https://prnt.sc/s9b1a0

Attached the full FS log from a session that reproduced this bug from a FS tester - Firestorm-1.log
This is the full log from where I took the "Error in log on FS EEP" image.

Looking at the log, I suspect in the case of the error occurring just after login that's it's a login race condition (when no region is set in time?), which would explain why the error message is random over many logins to the same region on the same viewer, without changing any EEP settings.

@sl-service-account
Copy link
Author

AndreyK ProductEngine commented at 2020-05-07T17:11:55Z, updated at 2020-05-07T22:05:03Z

For firestorm's log, source:
Firestorm viewer calls LLEnvironment::loadFromSettings during STATE_WORLD_INIT stage. It causes an asset request yet has no region to actually load asset from.

2020-04-10T17:31:52Z Startup state changing from STATE_LOGIN_PROCESS_RESPONSE to STATE_WORLD_INIT
2020-04-10T17:31:52Z LLEnvironment::loadFromSettings
2020-04-10T17:31:59Z Startup state changing from STATE_INVENTORY_SEND to STATE_MISC

Note:
In linden viewer the call was moved from environment initialization into STATE_MISC, where region is already present so issue shouldn't happen.

Potential scenario:

  • Set EnvironmentPersistAcrossLogin

  • Use any existing asset (do not edit, just use, or fixed one from world->environment)

  • Clear cache

  • Log in.

    Note: there might be a race condition.
    Should not be reproducible at linden viewer.

    My guess would be that firestorm missed couple eep commits. See commits named after SL-13029 (there was also one commit later in D500 under different jira, but that's a crashfix, not an asset issue).

@sl-service-account
Copy link
Author

Ansariel Hiller commented at 2020-05-07T20:17:16Z

FS EEP is up to date with the most recent LL release. I checked and LLEnvironment::loadFromSettings() isn't called before the call the LL viewer does in STATE_MISC in the build of the most current source. It might have been actually the case in builds before April 24th, caused by Firestorm's QuickPrefs instance being created early while initializing the toolbars.

@sl-service-account
Copy link
Author

AndreyK ProductEngine commented at 2020-05-07T22:04:45Z

Ok, thanks.
As I understand, in case of FS issue happened at login, but original reporter had a different issue that happened only when reporter switched to defined setting.

@sl-service-account
Copy link
Author

Dan Linden commented at 2020-05-09T00:17:53Z

Hi CeeEss,
Could you trigger that error again if possible, then before relaunching the viewer, zip up your viewer logs folder and attach it to this issue using More Actions -> Attach files.
This page tells you where to find the logs folder: http://community.secondlife.com/t5/English-Knowledge-Base/How-to-report-a-bug/ta-p/733545#Section_.3

Thank you,
Dan

@sl-service-account
Copy link
Author

CeeEss commented at 2020-05-12T18:27:36Z, updated at 2020-05-12T18:45:03Z

Hi Dan, cannot reproduce the error unfortunately as I was reading all the comments (no firestorm, default SL viewer).  To what I could gather tho in this environment coming to think of it, I was disconnected from the internet while logged into SL.

Nothing to note.  I have attached the cef_log zipped as from today.

 

 

@sl-service-account
Copy link
Author

CeeEss commented at 2020-05-14T06:20:04Z

Reviewing the cef logs I have discovered the inability of dataview to access SQL datasource might possibly due to corruption (cache??).

This is happening just now as I was pretty confident that the issue was fixed previously.

Might try to do a clean reinstallation of SL.

 

@sl-service-account
Copy link
Author

CeeEss commented at 2020-05-14T08:39:43Z

Clean reinstallation Second_Life_6_4_1_540593_x86_64_Setup.

New cef_log.zip for today as the rest of the local cache was deleted.

cef_log.zip

 

 

@sl-service-account
Copy link
Author

Whirly Fizzle commented at 2020-05-14T12:59:25Z

Hiya CeeEss,
Just to clarify, are you saying that you reproduced the "Could not load the settings for [NAME] from the database" error again on the LL viewer?

Also to add: it isn't the CEF logs that are needed.
The next time you reproduce the "Could not load the settings for [NAME] from the database" error on the LL viewer, log out and then before relaunching the viewer, zip up the logs folder at C:\Users\USERNAME\AppData\Roaming\SecondLife\logs & attach that logs.zip to this JIRA issue.

https://prnt.sc/sgos2o

@sl-service-account
Copy link
Author

CeeEss commented at 2020-05-15T15:13:29Z

Yea Whirly, cannot reproduce because the server somehow become unstuck while tried fiddling around with the settings.

cef_log? aha you mean zipping up the complete Logs folder... gotcha.

 

 

@sl-service-account
Copy link
Author

CeeEss commented at 2020-05-15T15:22:21Z, updated at 2020-05-15T15:29:16Z

logs.zip

There you go..

Ah Whirly, also to add... default SL viewer used.. no firestorm.

 

 

@sl-service-account
Copy link
Author

Dan Linden commented at 2020-05-15T18:28:04Z

Thank you CeeEss,

If you happen to get the error again and you are definitely connected to the internet, then please attach your log files again. I'll mark this as Needs More Info so it will allow you to add comments and files.

Dan

@sl-service-account
Copy link
Author

CeeEss commented at 2020-06-05T22:05:08Z

Recreated the error again.

Log attached..

logs.zip

 

@sl-service-account
Copy link
Author

AndreyK ProductEngine commented at 2020-06-06T10:40:17Z, updated at 2020-06-06T10:42:08Z

Thanks.

Relevant info from recent log:


2020-06-05T21:54:41Z WARNING # llcommon/llsdserialize.cpp(193) LLSDSerialize::deserialize : deserialize LLSD parse failure
2020-06-05T21:54:41Z WARNING #SETTINGS# newview/llsettingsvo.cpp(323) LLSettingsVOBase::onAssetDownloadComplete : Unable to create settings object.

Along with network failures (curl, packet_out_of_order, Http_404) and decode failures.

P.S. This probably should result in at least one retry.

@sl-service-account
Copy link
Author

CeeEss commented at 2020-06-06T13:10:13Z

 

Yes most probably the router is momentarily disconnecting from the internet during this session.

 

@sl-service-account
Copy link
Author

Dan Linden commented at 2020-06-08T17:16:01Z

Thank you CeeEss. Closing this bug.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

No branches or pull requests

1 participant