Closed sql-chris-albert closed 2 years ago
@sql-chris-albert hi! Can you do the following for us:
STG_EX_LOG_LEVEL
environment variable to debug
. You'll of course need to run storage explorer/reproduce the issue again after setting the environment variable. I see a whole bunch of these in the Dev console. Other than that nothing sticks out. Logs attached.
DevTools failed to load SourceMap: Could not load content for file:///C:/Program%20Files%20(x86)/Microsoft%20Azure%20Storage%20Explorer/resources/app/out/CloudHub/internal/ccs/ccs-0.1.0.js.map:
2021-08-10_115023_9528.log 2021-08-10_115023_main.log 2021-08-10_115023_renderer.log
Hi Chris, I am facing the same issue, with all the details you described, right after I have upgraded to v1.20.1. I removed the app and downloaded it again, same problem. By any chance, did you take this up to Microsoft? I would like to upvote there if you did.
Hi Chris, I am facing the same issue, with all the details you described, right after I have upgraded to v1.20.1. I removed the app and downloaded it again, same problem. By any chance, did you take this up to Microsoft? I would like to upvote there if you did.
FYI - I have uninstalled v1.20.1 and installed v.1.19.1 and it works fine.
@alexandru-pet Thanks for the tip. I installed 1.19.1 and everything is now working. Hopefully MS will be able to resolve the issue with the new version.
@alexandru-pet @sql-chris-albert
Albert's logs did not reveal anything to us about what is potentially going wrong. I've created a private build with additional logging. If y'all could download it, run it, gather the storage explorer logs (make sure to set that logs environment variable) and share the logs with us that'd be a big help.
I have two options for running the build:
Since y'all have downgraded to 1.19.1 for now, then you can use the zip to run the private build without losing your current install. Just download the zip, unzip, and then run the .exe in the root folder. If that doesn't work for you (and it may not, it's not a distribution method we normally use), then you can just go with the normal installer. Thought that will mean you need to reinstall 1.19.1 afterwards.
Thanks!
@MRayermannMSFT I downloaded the Zip and unzipped. When I execute the .exe it show a blank window for a second and then nothing. Have tried a few times.
@sql-chris-albert that's weird....I guess if you have time then to do the full install that would be appreciated. Unfortunately you will have to reinstall 1.19.1 afterwards. ☹
Awesome thanks! I can already see that something is "off" compared to what I see on my machine. I'll digest this and get back to you. Thank you thank you thank you a million times. :)
Ok I think those logs have helped me narrow down the area the problem is occurring in. I have a new private build with yet more logging:
Sorry it took me a while to get back to you @sql-chris-albert ...
So the latest round of logs give more insight into where Storage Explorer is not working right but I still do not have a lot of insight into why. And on top of that, the where that I have identified is one of the more opaque/black-box parts of our product.
So with all of that in mind I'd like to try something different today. This will be a bit more involved on your part, but if you have the time for it that would be helpful.
I've prepared a couple of private builds. These builds are from commits (aka changes) we made while developing version 1.20. I'd like to use these private builds to help us narrow down when the breaking change was made to Storage Explorer. I've looked through our commits for 1.20, and these ones are my current best guesses for what could be inadvertently causing this problem. What you need to do is simply install the build and find the oldest one with which this issue reproduces on your machine. I do not need you to get me any log files.
So here are the builds:
They are ordered oldest to newest. I suggest you trying them out in either of the following orders:
Let me know if you have any questions.
Also, @alexandru-pet if you are still around feel free to partake in this as well.
I started in the middle with 715c7ea5, and it worked fine. c5d38b93 worked fine as well. d2121ada is where the issue came back up.
@sql-chris-albert super! I will delve deeper into our changes from that time period!
Hey @sql-chris-albert (and @alexandru-pet), I have a new set of builds to try. Apologies to ask you to keep doing this, but since we cannot repro the issue and logging did not reveal the issue this seems to be the best course of action.
Right now I feel as if d2121ada
is what caused the issue. Out of all the possible changes from c5d38b93
to d2121ada
(inclusive) it is the one that comes the closest to touching what is failing to finish. So to verify that, I have a build of the commit before d2121ada
, along with a couple of others that occurred after c5d38b93
but before d2121ada
that are also suspect. Here they are:
The 2c2efa25
exe is the commit directly before d2121ada
, so I guess I'd recommend:
2c2efa25
, if issue doesn't repro then you're done/we've found the breaking changeb0eed2b6
bcce041d
or if it doesn't repro then try e08eb44f
Also, if you're curious, we're working on a range of changes from May 12th to June 22nd. So there's a lot of commits to sift through. I'm glad that we were able to so greatly reduce it last night though! The original range of changes was April 15th to June 22nd. 😬
Finally, if my guess about d2121ada
is right then we'll have to do some experiments to figure out why the heck that change is breaking the tree view on your machine. The change was just the removing of sending telemetry in an error handler which makes me feel like 🤷♂️.
Hi guys, sorry for not being able to join you on this investigation so far. I had some work related tasks for which I was highly dependant on the logs and couldn't risk losing access to them even for an hour. Today I am off early, but I will try to join from Monday.
Cool thanks for finding the time @alexandru-pet!
Ok @sql-chris-albert & @alexandru-pet, one more build to try:
With regards to Chris' findings:
d2121ada
being the cause was wrong.bcce041d
happened shortly after c5d38b93
, and this latest build is right before bcce041d
. Some additional questions (I have the answers to these from the log files but I want to cross check them with y'all):
41178ff2 worked.
I have not enabled proxy settings and my language is set to use system settings which is English.
Ok, I have a new build to try, and now we're back to sharing logs! I've decided to ignore the I can't fathom how that change broke the tree view init'ing and have added logging in the area of the change and also in additional areas that I didn't have before.
@sql-chris-albert & @alexandru-pet, when y'all have some free time please install, generate logs, gather logs, and then share logs from this build:
https://storageexplorerpublish.blob.core.windows.net/privatebuilds/4768-3/StorageExplorer-ia32.exe
Thanks.
PS: Chris, you left a comment on what looks to be Saturday! Please don't feel pressure to do stuff for this issue on the weekends (unless you really want to I guess). :)
@sql-chris-albert & @alexandru-pet
Ok. Still eluding us. I've gone crazy on today's build and have some additional things to gather/ask. First here's the build: https://storageexplorerpublish.blob.core.windows.net/privatebuilds/4768-4/StorageExplorer-ia32.exe
So, for this build:
CloudExplorer.App.
and then share what you find. Here's an example of what I mean:
Also some additional questions:
I'm out of the office tomorrow. Will get this done on Thursday.
Logs attached.
When I did the console search it didnt have anything for CloudExplorer.App. It did show some errors for just CloudExplorer so I included a screenshot of those.
I am not able to reproduce this on another machine at this time. I only use it on one other machine and it works fine there.
We are running TrendMicro for AV.
I do not have issues with any of my other applications. I do use Teams and VS Code and they work fine.
2021-08-27_094600_12808.log 2021-08-27_094600_main.log 2021-08-27_094600_renderer.log 2021-08-27_094600_se-markdown-panel_15084.log
Hi guys, jumping a bit in to let you know that in the meantime I got another laptop to use, where I installed directly the v1.20.1 and it worked with no issues. I guess it is pretty clear it is related to the machine used, because there were no changes applied on my account or on the connection string I used. Probably not related, but on the old laptop, where storage explorer did not work, I also had a weird issue with Teams, during calls and meetings, I would not see the camera of other attendees unless I would switch to full screen. Anyway, I have to handover my old laptop on which the issue occurred.
@MRayermannMSFT Any luck with the last round of logs?
@sql-chris-albert sorry, been busy getting ready to ship 1.21.0. No, the new logs have not been anymore helpful. The tree view is proving to be a black hole from which no information escapes...
ctrl+f
for cloud-explorer-panel
, find the first result and then...
<html
line, choose "Copy element", save that to a text file, and then attach that file here (see screenshot to see what I'm talking about)
C:/Users/calbert/AppData/Local/Programs/Microsoft Azure Storage Explorer/resources/app/out/CloudHub/html/CloudExplorer/index.html
, make sure your user account is able to open and read this file as well.This issue might be caused by us loading modules take longer than expected and timeout. For all of you who run into this problem, could you please try modifying a source file to override the default timeout parameter and see if mitigates the problem?
The file to modify:
C:\Program Files (x86)\Microsoft Azure Storage Explorer\resources\app\out\CloudHub\js\CloudExplorer\config.js
or C:\Users\yourname\AppData\Local\Programs\Microsoft Azure Storage Explorer\resources\app\out\CloudHub\js\CloudExplorer\config.js
/Applications/Microsoft Azure Storage Explorer.app/Contents/Resources/app/out/CloudHub/js/CloudExplorer/config.js
Somewhere at the end of config.js, add a comma and add one line below. Basically we are going to add a waitSeconds
parameter to the require
object on window
. This would extend the timeout period to 30 seconds instead of the default 7 seconds.
Before:
window.require = {
baseUrl: "../../",
paths: {
"ccs": "internal/ccs/ccs-0.1.0",
"es6-promise": "lib/es6-promise/es6-promise",
"jquery": "lib/jquery/dist/jquery",
"knockout": "lib/knockout/dist/knockout",
"knockout.mapping": "lib/knockout-mapping/knockout.mapping",
"underscore": "lib/underscore/underscore",
"underscore.string": "lib/underscore.string/dist/underscore.string",
"URIjs": "lib/uri.js/src",
"ActivityLog": "js/ActivityLog",
"CloudExplorer": "js/CloudExplorer",
"Common": "js/Common",
"Providers": "js/Providers",
"StorageExplorer": "js/StorageExplorer",
"text": "lib/text/text"
},
shim: {
jquery: {
exports: "$"
},
knockout: {
deps: ["jquery"],
exports: "ko"
},
"knockout.mapping": {
deps: ["jquery", "knockout"],
exports: "mapping"
}
}
};
After:
window.require = {
baseUrl: "../../",
paths: {
"ccs": "internal/ccs/ccs-0.1.0",
"es6-promise": "lib/es6-promise/es6-promise",
"jquery": "lib/jquery/dist/jquery",
"knockout": "lib/knockout/dist/knockout",
"knockout.mapping": "lib/knockout-mapping/knockout.mapping",
"underscore": "lib/underscore/underscore",
"underscore.string": "lib/underscore.string/dist/underscore.string",
"URIjs": "lib/uri.js/src",
"ActivityLog": "js/ActivityLog",
"CloudExplorer": "js/CloudExplorer",
"Common": "js/Common",
"Providers": "js/Providers",
"StorageExplorer": "js/StorageExplorer",
"text": "lib/text/text"
},
shim: {
jquery: {
exports: "$"
},
knockout: {
deps: ["jquery"],
exports: "ko"
},
"knockout.mapping": {
deps: ["jquery", "knockout"],
exports: "mapping"
}
},
waitSeconds: 30 // line added
};
@JasonYeMSFT That worked. Thanks so much!
Nice. We will evaluate the timeout parameter in our next milestone and extend it as needed.
We have shipped a patched version 1.22.1 which bumps all related timeout parameters to 30 seconds.
Preflight Checklist
Storage Explorer Version
1.20.1
Regression From
No response
Architecture
x64
Storage Explorer Build Number
20210720.10
Platform
Windows
OS Version
Windows 10 Pro 20H2
Bug Description
When I open Storage Explorer nothing is populated in either the account view or explorer view. To get my account to show up I have to click View --> Account Management.
Steps to Reproduce
Actual Experience
Expected Experience
I expected my accounts to be there in account view and storage accounts in explorer view.
Additional Context
This was working on this machine for a long time. Not sure what caused it to break. I have uninstalled\reinstalled several times, removed my account and added back in, and the behavior persists. I have another machine with storage explorer installed and I am able to use it with no issue.
At one point I was using storage explorer in the Azure Portal. I noticed a button that said open in explorer. I clicked that and it launched explorer locally and opened the folder. A message popped up saying the account had been hidden but did not give me any clue as to how that is resolved.