@ofer2 trying to post again.
Approximate time of incident: 2020-01-09 14:35 PST
Screenshot [if applicable]:
After 2nd Refresh:
502 errors and missing portals.
Also, I had great trouble submitting this post probably because it looks to similar to a previous one.
That's got nothing to do with Niantic. You need to uninstall whatever old Intel Helper app that you're using because it breaches Googles security requirements.
Cant see all portals now on Intel! 😪
@NianticOfer Is there any more intended testing today or tomorrow for Intel (given the last mentioned date is the 17th).
@Perringaiden we are preparing for a release of a new caching layer. Once that comes out, we will be transitioning back to the new server environment and testing will resume. I will let you know when that happens.
@NianticOfer i I don't know if this was expected or not, but on Sunday I ran my reference load and got 420908 portals. That's 200 portals lower than I expected.
Today I ran it again and got 42906, then reran it and got 42911 portals. While in exclusion, these three values could be accurate, with removal of portals and addition of new portals between each pass, the 200 portal drop seems aberrant.
Was there a significant removal of portals in Southern California on Friday/Saturday? Or is there a new issue?
Ahh, I was wondering why the intel map was moving so slow and not loading all portals earlier.
Hmm, Intel down, but we're past the 17th.
getting errors again whilst loading the map as well.
I'm completely unable to load tile key: 15_26147_13966_0_8_100
The response that's coming back is actually the NIA Takedown In Progress HTML page, but only for that tile.
For that location.
We havent changed Intel since I last posted, so I am fairly confident that this is not related to any Intel issues.
As I said above, we're working on moving the rest of the server, and we did an upgrade to the game servers to allow that yesterday (Intel was not changed) which seems to have broken some things. It was reverted, but we think we figured out its issue.
we are preparing for a release of a new caching layer.
I do hope it won't increase delays between ingame actions and their appearance on Intel.
@NianticOfer Any clarification as to why the Intel map is cudown?
@NianticBrian Intel are very slow now and dont show gray portals until i reload Intel. Please take look and fix it soon.
EDIT: Looks like the update at 11:10am PST has stopped these from failing. Not getting any errors on the load.
Hi folks, sorry for the late reply here but I dug into these reports with Intel Map and found the following: it looks like there was an issue server-side starting yesterday afternoon that caused some of these issues in addition to random failures, which most likely would have surfaced as increased ghost links and fields or specific portals getting into a bad state (might see the portal flip or not get attacked).
This was rolled back this morning and we're investigating further what went wrong but also how to avoid it for the next server update.
So, is the maintenance over? Because I have one very specific problem with intel. A new portal has appeared a few weeks ago ingame and it is nowhere to be seen on the intel map. It is simply not there and I have checked all the settings on my map. I have even asked other players to check their intel and they could not find it, but it is there ingame. How do we solve this?
What's the portal's location? At least roughly.
The coordinates are exactly 53.09375, 8.78216 and the name is Spielplatz. I can tell you this so on point because the local PoGo community is using some scan map which displays the pokestop according to that portal.
There something caching game state server side making Intel very wrong today. As I'm posting intel is showing the below field as up still, but in reality it has been down almost an hour.
This screenshot is from 30 mins after the scrolled back comm activity shown. That comm activity is Enl agent @sharp52 destroying and capturing all the spine portals in that field except the outermost layer. Even after clearing cache using different browser ( and a teammate doing the same on his system ) we can see the field as it was, still up. Clicking on the portals shows them as they were, not as they are. The only close to correct thing is the comm; even there the timestamps on those comm messages don't match the ones in my client's comm. Also the notification e-mails didn't come for all the damaged portals, only some. This is in Orange county in Southern California on 01-29-2020 from 4-5pm using desktop browser.
Since I know your working on revamping it I wanted to share the feedback. Please fix @NianticBrian
PS You guys might want to edit the links in the social drop down not to point to G+.
Here is a mobile screenshot too so its not device / browser / type limited.
Type of Issue: Other – State shown is stale
Approximate time of incident: 2020-01-29 19:45 UTC-6
Screenshot [if applicable]:
Other remarks: The blue fields anchored at portals north of the green crossing link were destroyed at 18:17 UTC-6 or earlier, which means intel is showing blue fields that have been destroyed for 1.5 hours and longer.
Are these also appearing in game?
Question is the difference between ghosting fields which appear on the map and the game and are an artefact of a failed dismantlement of large or layered fields but don't suggest an Intel issue, or fields that only appear on Intel which is an issue with the cache work they've been doing.
As regards my post a few minutes ago, by the time I managed to check what appeared in the scanner, the intel map had also cleared itself up. However…
Type of Issue: Missing portals
Approximate time of incident: 2020-01-29 20:17 UTC-6
I think that's a fairly clear-cut case of an issue with intel. With some repeated reloading (around 20:22), I did manage to get more of the portals to load, with some additional insight into what the browser is/was seeing:
Those requests in red are failing with HTTP 500 statuses and a “NIA takedown in progress” page.
@ofer2 should probably look at it then.
As far as portal flickering (which is what the above picture seems to be), thats a known bug which has existed for a long time.
As far as the caching issues, we deployed the new server for only a couple hours before discovering the issue and reverting the server back. We think we know what caused the issue, and so we will be trying again later. Thanks for the report!
As far as the caching issues, we deployed the new server for only a couple hours before discovering the issue and reverting the server back.
I just did a large load over Los Angeles, and came up 1000 portals short with 500 errors for a number of cells. Similar to January 22nd. I'll gather a list of the failing cells and then test them directly.