Something Going on With Anonymous Users?

[jonfroehlich]

@misaugstad pointed out to me tonight that there is something strange going on with the anonymous user contributions. Notice, for example, that the top ip address audited 15 street segments but contributed 4,381 labels. This is insane!

So, we need to investigate. Could it be:

• Vandalism or that some anonymous users have no idea what they are doing?
• A bug in the Scala/SQL query returning this data
• A bug in displaying this chart
• Other thing?

Who wants to investigate? @r-holland @sbower213 @adash12?

[manaswisaha]

@r-holland @sbower213 @adash12 Is anyone looking into this?

Since @sbower213 is now busy on a different issue, can one of you look into it?

[chishankar]

Yeah I can look into it

[chishankar]

From looking at it, my initial hypothesis is that the label count is just continuously adding the users label count to the overall total.

For example: 73.163.171.105 did 137 Audits and created 651 labels, but it added it to the previous total of 2020 and outputted 2671 as the total Label Count.

I will look further into this to confirm my hypothesis and fix this.

[jonfroehlich]

@chishankar your explanation doesn't make sense to me. :P

[chishankar]

I believe the user's total label count gets added to the overall total label count and that is why there is an inflation of numbers. I am looking into to it and will post a more comprehensive explanation once I know for sure what it is

[misaugstad]

After he talked to me in person, I think what he is saying is that... If you look take the bottom user in that table, it says they have 1796 labels. But if you go up to the user that is 2nd from the bottom, they have 1886 labels with 2 audits. He is saying that this user actually has 1886-1796=90 labels in the 2 audits. So the number of labels is cumulative, going up the table.

I think that he will soon discover that this is certainly not how this is calculated :)

[misaugstad]

Oh, and the above table is in the admin dashboard, in the "Users" tab, the 2nd table (the one of anonymous users). I then sorted by label count.

[sbower213]

I think I figured it out. One of the sql subqueries was pulling a lot of duplicates, and when the number of labels from each audit were added, it would keep adding duplicates. I added a "select distinct" into the query and think I've fixed it. I'll do some more cross referencing of the table to make sure I didn't omit anything, but this is what the new table looks like:

[jonfroehlich]

sweet. thanks. what's that weird 0:0:0:...1 ip address?

On Fri, Jun 30, 2017 at 3:40 PM, Steven Bower notifications@github.com wrote:

I think I figured it out. One of the sql subqueries was pulling a lot of duplicates, and when the number of labels from each audit were added, it would keep adding duplicates. I added a "select distinct" into the query and think I've fixed it. I'll do some more cross referencing of the table to make sure I didn't omit anything, but this is what the new table looks like: [image: image] https://user-images.githubusercontent.com/8651110/27751470-748a63de-5daa-11e7-957a-e8f63c2a4e31.png

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/ProjectSidewalk/SidewalkWebpage/issues/791#issuecomment-312356463, or mute the thread https://github.com/notifications/unsubscribe-auth/ABi-9R6FT6pAZDLraVhELcxHWpj6tlP1ks5sJU88gaJpZM4OKDvn .

-- Jon Froehlich Assistant Professor Computer Science University of Maryland, College Park http://www.cs.umd.edu/~jonf/ @jonfroehlich https://twitter.com/jonfroehlich - Twitter

[sbower213]

anything from localhost gets logged like that, according to stackoverflow

[jonfroehlich]

figured as much. thanks for confirming.

On Fri, Jun 30, 2017 at 4:22 PM, Steven Bower notifications@github.com wrote:

anything from localhost gets logged like that, according to stackoverflow https://stackoverflow.com/questions/10386875/host-ip-address-00000001-on-servlet

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ProjectSidewalk/SidewalkWebpage/issues/791#issuecomment-312364524, or mute the thread https://github.com/notifications/unsubscribe-auth/ABi-9Xk6RBbvtVDuF3PGytjtgtlN78Vlks5sJVj2gaJpZM4OKDvn .

-- Jon Froehlich Assistant Professor Computer Science University of Maryland, College Park http://www.cs.umd.edu/~jonf/ @jonfroehlich https://twitter.com/jonfroehlich - Twitter

[misaugstad]

Resolved via #799