[Intel Map] History inconsistency, sometimes missing

in Report a Bug
Intel data sends sometimes data without history data while its expected.
This leads to incorrect highlights for visited/captured/scout controlled
I get in most cases data such as
// getEntities response: .result.map["16_15940_11535_0_8_100"].gameEntities[...] [ "90b381c9b22a487e8d780626f55741f5.16", 1614009101125, [ "p", "E", 44804023, -687323, 5, 85, 8, "http://lh3.googleusercontent.com/GAkS1tBPFwP-uM8hXx656peqZ8OEyoWm_fvH3MusUZzTcM8ThOo_Tm1UQ9LDmBZP1QbFl8OFvtvLKTevHJLJB-qP9Q", "For\\u00eat du Bourgailh - parcours v\\u00e9lo", [], false, false, null, 1614009101125, null, null, null, null, 3 ] ],
but sometimes, history is missing
// getEntities response: .result.map["16_15940_11535_0_8_100"].gameEntities[...] [ "90b381c9b22a487e8d780626f55741f5.16", 1614009101125, [ "p", "E", 44804023, -687323, 5, 85, 8, "http://lh3.googleusercontent.com/GAkS1tBPFwP-uM8hXx656peqZ8OEyoWm_fvH3MusUZzTcM8ThOo_Tm1UQ9LDmBZP1QbFl8OFvtvLKTevHJLJB-qP9Q", "For\\u00eat du Bourgailh - parcours v\\u00e9lo", [], false, false, null, 1614009101125 ] ],
8
Comments
I think it's worth pinging @ofer2 here for fast and competent response. Could you take a look please?
I'm experiencing this when panning standard Intel map sometimes. It shows portals, for example, as uncaptured, and then if I refresh the page it highlights them correctly. This might be the underlying problem.
Although our API is free to change at any time, currently, as an optimization, if there is no portal history, we simply dont send the list of nulls. ("No portal history" means that it didnt find it, I'm not certain if this can happen due to timeout, but I dont think so)
This is on the same portal though, so it's not that no history exists. The second call didn't return the history that we know can't be reversed by player actions.
Oh sorry, I didnt see that the data was on the same portal guid. My main thoughts are either timeout, rate limit, or memcache error....I'll try and take a look and see if I can figure out where this might be happening in the morning.
Sorry about the delay, I got sidetracked on other work yesterday and didnt have time to look into this. I think I've found the source of this issue and should be fixed shortly.
Thanks for the update!
Hey there @ofer2 , just letting you know that the issue is still present, in case you thought it's resolved. If I open Intel in some place where I didn't open it within last few minutes, Portal History is often messed up. However if I refresh the page, it shows correctly for that same view. So this probably means that there's some problem when requesting history in a view that wasn't requested for some time (and wasn't cached?). I think if you open Intel yourself in a place where you know which portals you captured and which you didn't, you'll notice that.
@1valdis sorry about that. there are two issues here: 1. data wasnt being attached on a specific call and 2. data isnt present on first load. What I was referring to above was that issue #1 was fixed. Issue #2 is actually a more complicated issue having to do with mismatched caches. (One is caching on a larger size than the other which is why the second load works: because the second load will have the larger size filled). I've reraised the second issue to see if we can get some time allocated to fix it, but no promises. Thanks for bringing it up again.
@1valdis Just to follow up, I've put in a fix for this issue. The data on Intel Map should now match what you see in the scanner even on the first load. Please let me know if you see differently.
Checked, can confirm, looks great so far. Thank you, you're a legend!