SeAT - ❓-support - Page 5

ah i found something - so if i manually execute via console the "esicorporations" command with a character which corp is not in the list it will show the corp afterwards, but its not coming automatically

Seat\Eveapi\Jobs\Universe\Structures\Citadel anyone know how to close this job?

Is it possible to setup a squad that users can join if they choose but set it up so they get auto removed if they leave corp?

you can set filter in Manual type squad

Manual with no moderator with a corp filter

How to update prices in moons reporter? They seem to be wrong for me after migrating from 4.x to 5.x.

hi, my character\assets job use lots of runtime,

is it familiar?

ouch

that's probably why, it was spinning in place for a while because the supervisor had a bad morning after a reboot

Hey, quick question regarding Squad filters with skills and skill levels. So if i have a Squad for Freighters and i want to put in all Pilots who can fly a Freighter and have a Skill Level greater then 3 it seems confusing on how it works. So for example i build a filter:

• Match Kind "all"

• Match Kind "any"

• Skill: Amarr Freighers

• Skill: Caldari Freighters

• etc.

• Match Kind "all"

• Skill Level: > 3

This seems to put in all Pilots having any skill with a level greater than 3. how should it be done? Cheers

hell I would uphold a deny

after the amount of crying I just witnessed

just on general principle

(but I am not a nice person)

my rorqual mining days when you'd pop up in local way back when would agree with this

😄

to be fair

I only shot those who broke the rules!

OK, so no one knows nothing , thx

we know what SeAT can't do

as for what the CI folks in your alliance found out, well

that's for them to deal with

looks like you're trying to join a corp you joined back in june. If you linked any of the characters then, they'd likely have a record of it still. Even if it wasn't seat. I don't know if Frat use an auth for the alliance like other alliances do, but it's not unreasonable that a corp could gain details from the alliance auth. Just throwing out ideas that might send you in the right direction 🙂

that's a good tip

ITS not about frat / goons or else

problem 1 : not existing chars linked to my acount

problem 1: someone thinks those chars are yours

let's use correct language here

problem 2: not my chars linked to my account

yes, some security team members thing but this don't solve problem that i have.

and some of them are so that don't want discus this.

No one here can solve the problems mentioned. They have linked these characters together, they have decided they are yours. The problem is between you and the corp or alliance.

Q its : how ?

https://tenor.com/view/this-is-a-mystery-steve-blues-big-city-adventure-catching-blue-this-is-a-riddle-gif-27113480

I mean if you reach out to certain characters a lot or you have weird contacts ingame which look like you are keeping track of your alts or something one could assume those characters are yours. Also if you didn't want them to have the data should've never logged into seat honestly.

This link below will tell you if you have any active ESI to any Corp or Alliance, and that you can revoke those permission/access to your character.. https://community.eveonline.com/support/third-party-applications/

How you can track weird contact with char that was never created even game can find contact

you are really exhausting. Nobody can help you here and the best thing would be for you to talk to the HR team of your "desired" corp. To make it clear again: All information that SeAT processes is provided via various interfaces and is not only accessible to SeAT. How long an alliance keeps this data is not up to any of the people present here, nor is who links which data to whom! Nobody can help you here and you should be grateful for tips and ideas on where the data could have come from instead of abusing capslock. So go and ask your HR contact

Anyone got any ideas to this question maybe? https://discord.com/channels/821361165791133716/821361546608508938/1272128996792074241

i can do some tinkering on my server and get back to you if I find out the answer when I finish work in an hour 🙂

Would be awsome 🙂 thanks in advanced @user_106453105427640320

Yes. So we are aware of the problem and eventually want to fix it. The proper solution would be to combine skill type and level into one rule, but that is a slightly more complicated change than it seems, so we haven't got around to it. In the meantime, crypta added a mitigation that makes it work in the special case when the skill level and skill type are in the same group. In your case, that means you can make it work like this:

- Match Kind "any"
  - Match Kind "all"
    - Skill: Amarr Freighter
    - Skill Level: > 3
  - Match Kind "all"
    - Skill: Gallente Freighter
    - Skill Level: > 3
  ...etc

yeah thought so also, but either it seems it takes ages to get the users added to group since within the last 24h there still a lot of users missing. also i thought of adding more criteria and i did add it in the "all" blocks ... like "member of corp xyz" would you also do it that way?

It shouldn't take 24 hours. As for more criteria, why not nest the groups one level deeper?

all(
  corporation is foobar,
  character's favorite ship is corvette,
  any other rule,
  any(
    all(
      skill: Amarr Freighter
      skill level: 3
    ),
    all(
      skill: Minmatar Freighter
      skill level: 3
    )
  )
)

the UI doesn't allow the next nested branch you can only have 2 branches ... at least for me

actually, that might be the case. Sorry, I know the code behind squads better than the squads UI. Another thing to fix, because everything behind it is built for arbitrary depth group nesting

but yes, I guess you'll have to duplicate the corporation rules then

So yeah if the nesting goes down in depth that would be awesome. Yeah i've added the corp to each group which seems to work. Do you know out of the top of you head if there is a schedule which is responsible for the updating of the squads?

I think it is observer-based, so it doesn't have a dedicated schedule. it runs its check when a character gets updated

and it should do a full check for all members and non-members when you update the filters

hmm interesting, i look into that, maybe to much members for it to be fast 😄

Since we’re chatting about filters and such, I’d be curious of the relative performance impact of adding filters on assets or skill levels if anyone is familiar or has measured it. Thinking about adding some auto squads for tracking certain types of asset acquisition/loss or skill acquisition/loss or both but have some concerns about potentially impacting job performance if there are more than one or two of them.

Well thats what i was doing, getting groups setup with filter on skills and skill levels and it took ages to get the ppl into that squad

and the other thing im looking into atm is that ppl who delete the ESI Permission don't get kicked out of the squads

Somewhat an assumption as I’m not 100% but I don’t believe they would just from simply losing the token unless you have your environment set to delete characters with no valid token and also run the maintenance job after the token was lost.

Yeah thats an "issue" i guess since if someone invalidates his token for what ever reason - changing his password, leaving corp etc. there should be an automated process of losing all access until its back up working - this also should be the case for the discord access in my opinion

There two settings, one called clean up stale data, so users with out tokens get deleted..and another one called security level (though it might be part of the discord/ts connector) that can be set if any of their toons keys get deauthed they lose their access

by defaults its as long as any keys are still valid

IIRC the latter is called strict mode in the connector settings.

found the error, for what ever reason the server was not able to keep up with the requests so i stopped all process waited a little bit and restarted the jobs now it seems to work ... need to keep an eye out for it if it happes again

hello sir, i noticed mySeat\Eveapi\Jobs\Assets\Character\Assets job use lots of time to finished, how long this job time in your seat? its use about 500s around always.

hmm is there any way to put all users who are not correctly authenticated with a valid token into a squad. it seems "is not" doesn't work. as far as i could tell this is because the scopes are a "string list" so only contains seems to work but there is not doesn't contain option afaik

To what purpose? If its discord roles you can set them to strict and anyone without a valid token loses access

Just noticed when goint to /characters/ABCXYZ123/intel/summary I get DataTables warning: table id=character-top-journal-interactions - Ajax error. Any thoughts as to why? New install

to get a kind of reporting on which users do not have a valid token, since you cannot sort or filter this in the overview afaik - so i though you can put those users in a seperate squad to "sort" them

ah with ya

try YOURURL/corporations/YOURCORPID/tracking

you can turn off the valid tockens to see a list of invalid

it's a bit more manual

I'll see if I can get a squad to pull something automated my end 🙂

Just look at people who had roles.. and no longer but retains their ticker.

Or check your audit logs every so often

There are different ways to do it, but i just thought that it would be very nice to have some filter rule to make lives easier

Bare Metal issue : an issue where seat is unable to fetch or update from the server after a prolonged period of time. restarting the instance/vm does not fix this issue. If i delete the AOF file and reboot the vm then everything starts fetching properly again. Logs for Redis/PHP/Seat does not show anything noticeable. and the AOF does not take up too much space from what I see. Does anyone have any idea?

did you manage to solve this problem? I still have them, and I don't see any failed job by characters with this problem.

From the errors in the logs I see only this

but are there jobs that do succeed for these characters?

for example

the runtime looks a bit low for actual esi requests

hmm, Indeed, without paying attention to this, the rest of the work is completed in about 0.15 sec. for this character several jobs were performed in 0.01 sec

does the character use the correct scopes and not just publicData

full

indeed, for all characters with such a problem, work is completed in 0.00 sec and this problem only occurs with additionally linked characters, the main (user) does not have such a problem

I'm having the same issue with different characters within seat - looks like reauth them helps at least for a certain period

My thought on that is that there are too many requests to ESI for whatever reason and the token gets rate limited or something like this, i didn't have any other ideas

do you also have only alt? reauthorization helps for less than 24 hours 😦

what do you mean with "only alt"?

not main character in seat

it mixes through mains and alts

alright yea that worked. can confirm flipping off the email-as-unique-id setting fixed the issue

and yea if I can be blunt without getting in trouble, imo if you're willing to take over any kind of publicly used repository as the maintainer, then you prolly need to be ok with getting pinged about it and putting in maintenance dev time for as long as you are the maintainer

cause that's kinda the job lol

pinging with prs or critical issues ok....pinging any one no reason is no ok 🙂

yea of course

I haven't seen a load of more than 70% across all cores during operation in recent days. 4 virtual cores, 8GB of RAM

only seat from dockers run in vserver (cloud hetzner)

ubuntu 24.04 core 6.9.1

I'll try to increase the number of cores to 8 and see the result

huh, you run bucket updates every hour?

how many tokens do you have?

how can i watch this? this?

I generally list buckets and look at how many I have per bucket...

but that works too

but with an update every hour

and 30 buckets

you will have tokens updating every 30 hours

that's why you get the stale warning, I think...

+ it probably doesn't update at downtime, so it's every 31 hours

I would change the schedule to */30 * * * *

https://crontab.guru/#*/30_*_*_*_*

How can I do this in Seat? In the interval settings in the web interface I only see these options

seat docker

I learned that you don't have to choose, but specify it yourself, thanks for the advice, I'll be watching

cool

found out that cron can't be executed every 31 hours 🙂 I'll think about what to do, probably I'll leave it once every 2 days

I don't understand 🙂

all your tokens are divided into 30 buckets

seat:buckets:update triggers the ESI update of a new bucket

by default it runs every two minutes

so it should be run more often than every hour if you want all tokens to update inside 24 hours

oh, I misunderstood you 🙂 thanks

@user_334234098195234817hi, how long about your Seat\Eveapi\Jobs\Assets\Character\Assets jobs use? in my seat, someone use about 500+ seconds.

This happens sometimes, but not all the time

for example 🙂

yea, i have same fail job with you, and how do you setting about Attempts, i never see 89 times

I have no more than 10 such mistakes per day, and there are also successful works, so for now I just don’t pay attention

didn't help 😦 I was monitoring a previously added token that hadn't been updated for a long time and asked to update the token on one character information on these characters is still not updated. @user_131158012831203328 any other ideas?

every 30 minutes there is still this error

That endpoint has an issue where if one id of 1000 is not okay, the whole request fails

I can ignore this problem?

mostly yes

ok

no idea

so i have SeAT installed as bare metel. I want to run my own pathfinder instance but this is a docker image. Can i run both and route the traffic correctly ? im asuming i will need to configure traefic

yea

You can but it would be just easier to move PathFinder to a different port and do not mix configurations

Put them both on odd ports and run nginx, caddy or similar in front of them with acme cert etc

Structure alerts via discord integrations seems to fail every time, but the webhook itself is valid and works ?

Maybe I'm missing something simple, if so sorry just confused tbh

does anyone know how to fix the Ajax error when searching for assets?

it looks like a bug. The test notification is just text. Most likely, something is wrong with the message we send to discord, as most notifications use some fancy formatting

Ah ok so less of my fault... that makes me feel better. Almost rage re-installed

Just an FYI, you'll probably want to delete the first image and/or censor the webhook URL. Probably best to delete and regenerate it. Not that it's necessarily a major issue, but could allow someone to send messages to your channel if that's your whole webhook URL.

This hook is invalid at the time of the screenshot 🙂

Made a test one to ask for help, thanks guys sorry to bother.

Just checking to make sure 🙂

That looks like a bug. Here is how I usually approach such problems: 1) Debug it and figure out what is wrong 2) Devise a fix and implement it 3) Share the fix with everyone else if the error is because a seat ajax endpoint fails, you should see it in your logs

no logs are spitted for redis, seat, php

I mean I record all discord traffic including images with a proxy for archiving. Deleting is not enough

Since it happens in the browser, maybe open the browser dev tools and look in the console and network tab

debug mode shows nothing as the only thing that is hinted is timeout, i increased the nginx timeout/ and php but it doesnt seem to help

Oh I meant delete the webhook on Discord which which could have been clearer for sure, but yes, absolutely.

the console for the webrowser only shows POST 504 (Gateway timeout)

Do you have a absurd amount of assets? it work for me, but I can imagine it hits some timeouts when the DB query takes too long. I'm pretty sure we have no optimizations/indexes in place for searching, so it is rather inefficient.

i guess we have an huge database with a lot of stuff. is there a roundabout way to get this to work? reindexing?

let's move this to a thread in #channel_1113334797201317941 since it might take a while+loads of discussions

👍

i create then?

thread is up

Was the attacker not in an Alliance? If so it's fixed by this but it still hasn't been merged. https://github.com/eveseat/notifications/pull/101

So then it is that. Unfortunately, we can't do anything until @user_301981661761896449 is back and merges and releases the fix

yeah we are a bit understaffed in the development team, and only really one person (crypta) can merge and release things

(You can run a Developer install and checkout my branch. But that is "Advanced")

even if pathfinder uses traefik ?

well that's up to you eh?

lol. sorry ime pretty new at this and not sure what to do

right

so im using bare metal seat running on port 80 and 443. I installed pathfinder via docker to ports 81 and 444 . but its not secured. unsure on how to route traffic ?

Since we are messing with the new structures alerts on the notifications modules maybe the development team can include the old starbase alerts (fuel, attack etc etc) on the already existing categories like you guys done with the new structures?

like including the POS's on the structures categories

and avoid messing too much with the code

eventually it will be usefull too those wh corps who have a bunch of POS's

its an idea

or even look at the code to reactivate it, its still there but deprecated since 4.0

is there a manual way (SQL query?) to clean up your asset database from characters that were deleted from seat. I want to do some cleanup and from what I was told, seat doesn't really "clean up" the backend when a character is deleted.

if you delete the toon, all the data goes away

Seat\Eseye\Exceptions\RequestFailedException: Character has been deleted! in /var/www/seat/vendor/eveseat/eseye/src/Fetchers/Fetcher.php:344 deleted character can be filter?

Hello! I was wondering if someone could help me the SeAT billing plugin. I have some members who have paid their bill (correct code input multiple times) but it isnt registering that they have done so. Is there a way to edit or remove a bill so that the person does constantly see that they havent paid one

#channel_821361566184112181

Ty much appreciated

on https://eveseat.github.io/docs/upgrading/from_seat_4_0/docker/#update-the-env-file section, you update the seat 5.x value with the old value of the variable yes? example: REDISHOST=redis from seat4 is now: REDISHOST=redis new .env file for seat5 is REDIS_HOST=cache, but you change that yes?

i do see this in the example .env file

# When using the redis driver, this is the name of the redis connection. If using 'cache', it will reuse the existing
# redis cache configuration found in seat/config/cache.php. It must be a configured connection

but i guess it still seems to me REDIS_HOST should=redis, if seat4 .env was =redis

didn't seem to be doing anything so killed upgrade reverted snapshot and giving =cache a try

ya is progressing now

concerning error though cannot create cache directory

I was fairly sure that was just a table of things that would need to be updated and the original values, but you were quicker than I could double check my .env to confirm. Fairly sure it’s cache though 🙂 I’ve seen the message you’re likely referring to, and if it makes you feel any better I haven’t seen any issue related to that myself.

ya gonna check another seat5 i already have running see what that looks like

ya cache on my other seat5, i missed appkey though, restoring snapshot going again lol

for future others searching REDISHOST, it should be REDISHOST=cache

even though old value is REDIS_HOST=redis

hmm, in my other seat5 that's working it's missing the whole ESEYE Configuration Section

error that included the

Use the option --with-all-dependencies

seems to be resolved by double checking all plugins desired had seat5 versions, and checking seat5 name was correct, seat-fittings for example, changed from dengarr to cryptatech

doing the final down/up but yeash, that was so much easier this 2nd upgrade i did compared to when I did my personal one when 5 first went official lol

Hello, if I migrate the seat to another server, is it sufficient to copy the .env file and the database? Or are there other things I need to consider?

if you are using seat-info and have resources uploaded, you also need to move them. otherwise DB and .env should be fine

cool, ty 🙂

Hey. How much trouble will I face if I change the developer account? Will all users need to log in again? One of our members who originally managed SeAT has switched alliances.

Everyone will have to reauthenticate

inventory:assets is it possible to slow down this cron job? it's currently every 10 minutes, we'd like to slow it down just a bit, it doesn't show up in available commands, or will it not until I delete the current job?

If you know they won't deauth the app you can just continue as normal.

It's not like they can just view stuff on seat just because they retain the dev app stuff

I'm assuming he'll do it sooner or later, so I'll just let it run for now. Thanks for the info.

Question.. I have player that "registered" to my corp seat, however, when i look him up the status say "unknown" on his character sheet... does that mean he registered and then revoke his ESI? and if so, is there a way i can verify it?

Check his token. Also server need some time to pull his data its not instant. It depends how fast your server is as well. How fast did u check that character?

Also check logs if there is no issues with that char.

didnt pull since 8 hours.. only after i clicked on "update" then it started to pull the data..

im running 4 cores 12 gb of ram (docker vm). usally its quick..

how many tokens, and what's your bucket update set to?

around 100 tokens. and the bucket is set to seat:buckets:update

  */2 * * * *

shouldn't take 8 hours then

The uptime of my VM was only 6 days.. so i just rebooted it.. ill see want happens.. thanks

Yeah, do you have monitoring of cpu/memory use?

I don't.. Should really get Grafana..

Do i understand it right, that the Discord PlugIn only works, if the Members identifys there Discord Account over the Connector -> Identities -> Join Server? If yes, is it possible to set the Viewing right for the Connector -> Identities for the default user?

You need to make a squad that pulls any user that logs in and gives them access rights

Let me double check my settings so I give you correct info on what works for me

I have a squad with the "auto" filter scopes is not publicdata. This seems to grab everyone as soon as they make an account Then you make an access management role that gives view permissions to connector then you go back to the squad and add the role to the roles bit... that should do it

There is the Everyone role as well that you get once you've authed. I believe that comes with SeAT

hmm

Scratch that. Thats a custom one

well i dont see that window anywhere...but i did create a squad on auto and attached a role to it with the rights of the Connector...and its slowly filling in the Members. So...looks good atm

If it works it works

Hey, anyone got an idea why the moon reporter calculates the Refined Materials wrong - here is an example from janice vs seat

Hmm? What is wrong? Quantities and volumes seem to match each other?

Prices aren't matching up

seat uses esi's adjusted_price, janice the jita market

was that always this way? since the upgrade von v4 to v5 this occured. or is there a way to change the calculation to janice within the reporter?

Hm...what cronjob needs to Run to automatically sync the Discord Users with the SeAT Plug in? 🤔 Atm the bot only Sets roles when i manuallay start the sync. And the cronjobs that are there are a bit confusion Name wise...

on my side note.. not sure why the assests are not showing up. even the job did complete for the token.. either it didnt work, or this person doenst have any assets or isk...

https://github.com/zenobio93/seat-connector?tab=readme-ov-file#commands seat-connector:apply:policies would likely be what you're interested in setting on a schedule.

Hm...thats the one we are using....well must be an Error on our Site then.

is the "null" normal?

What's your cron expression and are there any job errors related?

Sync sets + apply policies

Add those to the schedule

I dont understand. Some of the tokens have this error message. I checked the job and it shows it completed without errors....

Have you adjusted any schedules or removed any? Is your SeAT install up to date? Any errors in the logs?

i did but a long time ago i changed it..

esi:update:status     0 * * * * 
eve:update:status     0 * * * *  
inventory:assets     0 * * * *

I think from 1 min to 15 mintues..

everything else is standard..

uh

not sure what you did

or I mean

not sure why you did that

but IIRC these really should be left like this:

esi:update:status     * * * * *      
eve:update:status     * * * * * 

i changed it because there was a high DNS queries.. over 10k hits.. wanted it to cut it down a bit..

...what

I think i found the logs..

I believe the status items are used to ensure things are reachable and thus are a prerequisite for other jobs, would be best to revert them.

yeah 100%

regarding my message above https://discord.com/channels/821361165791133716/821361546608508938/1275816766248845394 could it be that this changed with seat v5?

because a corp mate said that the prices where correct before we did the upgrade

If you wish to reduce DNS queries it would likely be better to consider other options like DNS caching locally rather than modifying the schedule for these particular items.

yeah that's the way

(I have never cared about dns queries)

im not sure where and how the dns query count are comming from but eve itself got a 60 sec cache in place and the "gateway" or dns servers should be able handle a couple of thousand queries a day without any issue at all

Would be curious to hear if that takes care of it, but here's a reference to why at least eve:update:status is important. https://github.com/eveseat/eveapi/pull/401

looks like it did... the token status are all green..

Ok.. looking at recursivetree notes, it looks like 15 mins was too long of wait... it is more of a heartbeat check..

I'm not sure if this changed myself, but what is your setting for market price region in SeAT Settings?

The Forge - you ment the "Market Prices Region" right?

Yep, was just curious how you had it set. recursive_tree + crypta have plugins for price providers like Janice but that's only available to other plugins, at least to my personal awareness. I don't believe there's a current way to use an external price provider for built-in stuff.

Yeah we got that installed, but its like you said that it doesn't is configurable for built-in stuff afaik. But what wonders me is, what changed from seat v4 to v5 that the prices suddently changed

the source, more than likely? 😛

sounds like a setting change

😄 yeah but i didn't find something within the code, that said im by far not realy capable of understanding everything there.

any of u guys having issues auth'ing into your SeAT? I'm guessing it's this but just checking

looks like there is

I mean login services are being restarted..

So SSO/Ingame actions and logging in are affected including ESI

@user_164235653008064513 re your PM - I have a 4000+ token SeAT instance on a VM, bare metal install. 18GB ram, 4 cores of a Ryzen 5700G on proxmox, mirrored enterprise SSD storage (ZFS), with the following SeAT tweaks:

• keydb as a replacement for redis (not sure how much that really helps)

• bucket update is set to every 20 minutes ( seat:buckets:update */20 * * * * )

• redis eseye cache

• a cronjob that nukes the seat cache (redis) every 24 hours

crontab -u www-data -l
* * * * * php /var/www/seat/artisan schedule:run >> /dev/null 2>&1
55 23 * * * cd /var/www/seat && php artisan cache:clear  >> /dev/null 2>&1

Also used to run a SLOG device on a 32GB Optane drive, but with several hundred GB of sync writes per day, it was in danger of being worn out in <6 months, so I am currently running without SLOG on ZFS. With NVMe SLOG, CPU load was lower and the storage drives were way less taxed

hey, thanks for responding, I was thinking about brining this up on the weekend . I'll look into keydb that one is an interesting one ty.

bucket update is the biggest change

to get load down

it does mean your updates are less frequent

but with 30 buckets and 20 minute refreshes, you get 72 refreshes per day (ca)

or about 2.4 full token pulls

the bucket update you mentioned is the seat:buckets:update */2 * * * * in the scheduler?

yep

I set that to seat:buckets:update */20 * * * *

some one help me ，my i18n does‘t work

ok, the discord bot setting the right ticker, etc... but it does not assing roles, what do i have to do here ?

and how do i assign seat to group alliance members to the alliance member role, etc ?

did you read the docs?

i installed it regarding to docs yes

ahh wait

hmm

does not work for me

it does not autogroup members regarding to their corp

do i need a plugin for that ?

No @user_283311733060272128 , you go to the connector -> Access Mamagement und set the filter accordingly to the Seat-Roles - you need to have the discord roles created before that so that the "Set" is synced

So for example:

You create a Discord-Role "Member" then go to the connector and sync the sets "Update Sets" in the Connector Settings

Then you go to the Connector Access Management and create a specific filter: Like -> Corporation Filters -> Corporation (Choose your Corp) -> Set "Member"

once this is done it will sync the user roles

works now

but how do i group my members by alliance ? and give them the appropiate role on seat ?

atm i do this handish

via squads

squads have filters based on characers / account

and those can be connected to the other services like discord

but i have to add them manually ?

you can also do it in access management directly

nono , discord roles works fine now

thats not ht eproblem

the problem is that i can group registered members by alliance name and give them the proper roles

atm i have to do this manually

true forgot about that, using squads for ages for this purpose ... better not mix same things with different solutions 😄

thats what we both where referencing too

unless you mean roles on seat for example. In which case @user_150663988638646272 is probably spot on with the squad setup solution. Setup an alliance squad, and give that squad the roles on seat you want. then set the filter to auto alliance is {alliance name}

but where do i set my alliance id, that seat knows who set to which group ?

in squads make a new squad, name it after the alliance

then set the filter like I have mine

it'll pull everyone from that specific alliance into the squad

then you can make roles under settings, access management and set those roles to apply to the squad

So you basically do not set any alliance id, you filter by name

i'm still trying to figure out how to lower the chances of the AJAX 504 timeout error. We always get this more often than not when our members try to use the character asset search.

can't really help you there

Regarding to my question above, is there a way to get the moon reporter prices use janice or some other provider instead of the adjusted prices? any ideas on that?

I deleted a shitload of toons from our SeAT installation. But for some reason they are still listed in plugins etc. - I have Cleanup Stale Data set to YES. Can that be forced somehow?

Which plugins? They may hold their own tables that seat itself does not maintain via cleanup.

Fittings plugin specifically

I'm not familiar with how that plugin handles, but do you still see the character otherwise? If so, do you have the seat:admin:maintenance job scheduled? Stale data is only cleared when that job runs.

You can’t really remove data. Seat might have a few features like remove stale data, but they don’t don’t do a lot.

From a technical view, it is not impossible to completely remove characters/users, but there is just no attention spent on it during development

If you don’t need a toon anymore, remove the token so at least no more data is pulled, but going much further doesn’t make sense since it doesn’t really work anyways

Would a reinstall of the plugin pull only present data

Or would it have it cached as well?

Actually not present as a schedule.

Might try adding it

The job also does some table optimization that can be helpful for performance and some other cleanups, even if you don't have the stale data option on (which I don't because I want to keep the historical knowledge) https://github.com/eveseat/eveapi/blob/5.0.x/src/Jobs/Maintenance.php

reinstalling a plugin doesn't delete its database

So SQL crawling is the only way to get rid of data. And that in itself is a no-no right?

It depends if you know what you are doing. Personally, I feel like I know the seat db well enough to understand the implications of deleting this and that, but if you don’t feel confident better don’t.

Another question: why do you even want to delete the data in the first place?

We wont be needing it anymore and it "corrupts" the pulls we're doing from plugins etc.

Fitting plugin includes all the toons ever listed and we had a lot of those. 1000+

Så the report page for doctrines are pages long and difficult to read.

you got lucky on timing 😄

I was looking at discord and saw some traffic

hehe

but I appreciate it ❤️

I think their issue was with asset jobs taking a while and timing out, it’s something I encountered myself on a particularly gnarly user with tons of assets after the fixes went in for the assets disappearing every other run. I need to look at indexing assets at some point but I have to figure out what the SQL queries are doing to determine what good indexes might be

if someone wants to take a stab at optimising the SeAT SQL stuff, I will pay for many beers

Yeah, I have a feeling it’s looking by ID or something to find the asset to update it so it might just be as simple as a single field index. Just have to find time to dig into it.

theres also the ESI bug that if you have enough assets it just will never return anything 🙂

woah

so that's where all my trillions went

😛

( /s )

does this affect personal or corp assets

Are you talking about an actual ESI bug or a situation where the assets job for the token runs so long it times out? Not seen a failure of assets in a long time personally, but would be curious what you mean by “enough assets” and whether it’s a logged defect.

an actual bug

if a corp or a person have more than x assets in a single location it breaks the asset end point for that entity

Oof

so you guys with 4000+ tokens don't get much ajax error correct? and i'm also assuming you guys don't use cloudflare or w/e

Even if you use cloudflare it's just going cache the frontend part and not help with backend data processing.

The root cause is the call doesn’t return in time - if you want to fix that you either need to figure out how to get the call to come back within the time limit or adjust the limit on whichever proxy or process has the timeout.

Eh, some. I don't generally care too much about those

I just activated Discord Notifications and certain ones for example war declarations it sends multiple for the same event, about 70 of them. I use the docker version. Any ideas how to solve that?

do you have multiple characters as notification sources?

I have our corp set as source. So it would be better to have 1 character for that kind of notification then?

Oof

So did I get that right? I get a notification on Discord for every toon in my corp that gets that notification in game?

Seat doesn’t do any “de-duplication” for notifications, so any character that meets the criteria that receives the notification will trigger a notification. I usually use a CEO or director - basically someone who is unlikely to rip their token, as the scope.

Yup, I just figured it out. But I guess it depends on what notification it is.

Thanks

Yeah, they’ll differ based on the type of notification, like account creation for seat isn’t based on receiving a notification in game from ESI. I just have a couple different notification groups for the in game stuff.

Makes sense

Oof, but understandable

For aD we polled keys in a rotation to get quick updates

It would be nice to have aggregation since tokens are handled at different times and could result in a faster trigger or “redundancy” of important notifications, but that’s how it works today. I’m sure the devs wouldn’t mind a PR for a system to handle it if you have the skills. I certainly don’t. You can also widen the scope for important stuff and deal with however many messages you get if you prefer expediting the potential.

I don't either, a buddy wrote that piece of software 🙂

I just noticed that the SeAT DB is reading with 200MB/s from my drive for 1 hour straight i do not see any increased load on the Horizon Dashbord so i have no idea where this is comming from the only thing in the schedule that i see is the esi:update:prices that would start at exactly the time i see the drives going up. is that normal behaviour?

where can I see or get the root password in a docker container?

I use this instruction, but I get under the user seat although the screenshot shows that the login will be under root 🙂

Any ideas on why searching for something in the global search at the top of the Seat page and opening an Evemail from there goesn't work? It just sits at loading. I can open the Evemail if I go to the character page though.

you dont need it, the user the container is running under will have all the access it needs

I think there may be a bug if the character isn’t able to be looked up where it fails (in any of the from, to, cc fields). Haven’t confirmed myself but have seen similar behavior.

My guess would be biomassed characters but haven’t looked into it very far

encountered this problem https://discord.com/channels/821361165791133716/821361546608508938/1248802875606437928, and I also want to give permission to edit the folder in the container

I think the fix to this problem might be that not all tokens are updated

It's very interesting, if I log in here under a character who has no problems, then I log in normally, but if I log in under a character with problems, then I get a white screen. Checked in different browsers

Now the screen is white and there are no problems when logging in with a character I also checked it on several browsers 🙂 and even tried it through a VPN

It seems like I'm under some kind of curse 🙂

welcome to esi voodoo

honestly ... if things are so fucky

wipe all the weird tokens

add a character to the account via the button Link Character. Token information is green. Only public information is updated, although the scope is full. After about 24 hours, the token icon turns yellow and no new information about the character appears during this period.

this does not help, deleting the token even on the CCP website (https://community.eveonline.com/support/third-party-applications/) and reconnecting the character leads to the same result

Hmm, there was a fixed issue a while back where linked characters were getting the wrong scope applied at one point but I can’t recall if that was the behavior. If you’re up to date you can try logging in as the primary character that owns the account for that person, then link the character having the problem again and see if it resolves? If that doesn’t work, if you delete the character does it then fix itself?

okay... so here's a stupid question: how many valid tokens do you have at this point?

It almost sounds like your characters only have public data scopes based on the behavior

The ones with the “problem”

hmm, how the f do I check how many mains I have registered again? :v

how can I see this? in general I can quickly tell about the corporation

I don't remember that's why I ask 😄

if you look at the corporation, then out of 276 tokens, approximately 35-40 percent are yellow tokens, the rest are green

mmmh

a execution time of 0 to 0.1 seconds hints at that no esi call is made in the first place. There can be a variety of reasons, this is not complete

• no error limit budget remaining

• insufficient scopes

• another job of that kind already running

• request is cached

• and much more

I've turned on the debug log level now, I'll be watching

just don't leave that on though. I once let it on on my dev instance overnight and got a 10GB log file

a record appeared on a character with a token problem Debug 2024-08-30 19:01:51 local Loaded cached response for get -> https://esi.evetech.net/v5/characters/1322297279/?datasource=tranquility Log index 13785. Click the button to copy link to this log entry. 13 785 Loaded cached response for get -> https://esi.evetech.net/v5/characters/1322297279/?datasource=tranquility Context: [ "32ff4da10a77cf0e27ec2d7114cac3863bcec43f9320bf1bdfaec219"

gets character information from cache and doesn't try to update it?

lovely

[Jobs][Middlewares][7b0937a5-9eb0-4e97-b7af-5225965ea8f2] Check Token Scope -> Removing job due to required scopes not matching with token related scopes profile. Context: { "fqcn": "Seat\\Eveapi\\Jobs\\Assets\\Character\\Locations", "characterid": 1322297279, "scopesprofile": 0 }

hmmmm that's ... something?

Laravel logs

by character with token problem

I don't have enough knowledge to understand what this reason for deleting a job

Sounds like the job is needing an ESI scope that the current authed token doesn't allow for

I'd blow all the tokens away and have people re-reg

if this is a recurring issue

looks like there is no other way out 😦

Do you have multiple SSO scope profiles?

did you set it up to not ask for all the scopes at first?

You can find under Settings -> Single Sign-on

yes, two profiles, one is just public, the other is full

remove the "public" profile

Now hang on lol

Don't just start deleting stuff

I assume you're using these two scope profiles? Which is the default?

lmao if the default is publicdata

yes indeed, the default now is that public... it didn't bother me before, but at what point did it become a problem

This fix was put in place as well which was causing some issues with multi-scope https://github.com/eveseat/web/pull/678

So it probably doesn't matter necessarily that the default is public if you have a selection, but if the main character on the account is "public" scope, it will use that scope for linked character additions

now I switched to the default full key do I need to relink all the characters now?

Well, so there's more to it than what the default is... were you using a specific SSO button to sign in before on that page?

For example, if you were using the customize sign in page option

the main character all log in on button with full access when linking to my new character I saw a request for full access

So there are two buttons, one is for public SSO scope and one is for full access scope - and users are linking additional characters to their main or are they logging in with each character individually?

I'm now beginning to understand that this problem probably arose after the update to the 5th seat. Since the problems only affect those characters who linked them after the update to version 5

to add non main characters use Link Character

Could be yeah. So the big thing here would be the scopes that the "main character" (the one that "owns" the account on the Settings -> Users page) is setup for. If that main character is configured with public scope, any other character they link should also try to use that scope. If the main character is full scope, it should (hopefully) be full scope.

You can see which scopes are authorized by clicking on the user account from that page and then clicking "Scopes" next to the character.

sry google translate

The main character is registered via the full access button when submitting applications. And there were no problems with the main characters

So if you find one of the characters having a problem in user page, click on scope, does it have full scopes?

yes, it was checked, such characters full scopes

For example, for my character who has a problem There are more entries below, the screenshot shows what was on the screen

I just updated the token of my character who had problems, the token turned green. I'll watch. Let's see what happens in 24 hours

my main character was registered before it became possible to register only with a public key

yeah

Interesting... yeah, I am not sure then. You could certainly do what Wibla suggested and just delete and re-link - but it might be better to find a proper solution if you don't want to lose any data, would be up to you on how important it is I think. Further troubleshooting probably needs someone with more knowledge on if an upgrade would change the behavior or cause issues with how the linked scope is determined - but possibly changing the default and logging in again might fix it for those older characters. Not sure if this would then cause a problem if you wanted to link characters with public scopes later though.

It looks like setting the default full key and updating the token gave a result!

Well, that's good news at least. Although I wonder if this breaks linking more characters to a public profile then, for you.

the character with the unupdated token continues to have errors, but mine is ok

I will monitor those who received only the public key, but this will take 24 hours.

You may also wish to add a new character with public scopes only and test what happens when you link a new character to that main - if it asks for full permissions or not.

This assumes that you care about the public SSO profile though 🙂

1. I create a new user with a public key 2. I try to add the same character via Link character 3. I check if the full data is requested right?

Yes, basically wondering if the system works as expected for your public keys now or if you will have problems with those after making the change.

requests

That seems unexpected. The original character only got public scope?

yep

If seat web version is up to date, it would require someone smarter than me to understand what is causing it.

number of users in the configuration/users panel

yeah I got around to look at that later 😄

haha

my bad

no bad, just took a bit...

Thank all very much! I hope that after 24 hours we will see that everything works 🙂

🙂

the time to complete the job is now normal more than 0.01 seconds for those characters who previously had problems with tokens and the tokens were updated

anyone got tips on how to mod the styling for SEAT? I attempted using the provided docs, but im not sure if its out of date or what but I cant seem to get it working.

https://discord.com/channels/821361165791133716/821361505735278592/1235117945693143070

thats some one thats sorta rebuilt all the css for it

its pretty goiod

characters with updated keys have their data updated! A huge thank you all again!

I have a problem with the notifications atm. SeAT is sending sometimes Notifications 2-4 times. Killmails for example are getting always send double...

did you add multiple characters as notification source?

Nope. One Entry in Integration, one in Alters and one in Affiliations

hello, my instance of seat seems to be trying to pull every corp and alliance possible ingame attempting to load their stats causing a huge bog down on my server. How can I prevent from mass pulling every corp/alliance ingame and only the ones added into the system? this has been solved, thank you

Usually, there are filter you can add to something.. see if you can only filter to your corp.

i still have an servererror 500 after authorizing, but webpage works if i visit the origial url of the seat server without the token

what is that for an error ?

some service not running ?

Still have the problem with the notifications atm. SeAT is still sending sometimes Notifications 2-4 times. Killmails for example

Where are you pulling the notifications from. If corp notifications for example, it'll pull for each instance of persons who receive it. So 4 director characters will send 4 director level notifications

In those instances, pulling from a single character should mitigate duplication

Anyone got an idea why some users got this message with the SRP module

When i do this as admin it works, but as the users who lost those items it doesn't work

when i add it as admin delete it and retry to add it as imporsonated user it works also

are the moon drills implemented yet ?

for seat notifications ?

Could someone please post a screenshot of the default job scheduler? And thank you in advance.

I sometimes have the same problem, but it is most likely related to the ESI CCP.

Keep getting failed jobs... I also get sometime 2 notification message of the same message or sometimes 10 of the same messages....

Yes there are.. Go to the "notification" to set them up for discord, email or slack..

Is this the regular moon drills or does it include the metanox ?

@user_1202730251763843082 moon dills yes, don't know about metanox, try it and let us know.

well for the killmails for example, i pull if over the Corp Affiliation. because i want all the corp Killmails. But thats only that one entry

Metanox also works

Ty

got like 17 notifications send to discord yesterday for the same shit. And it can't be related towards Directors in the corp. There are only 2 atm. The problem should be something different. Also how should i even pull certain Corp related stuff from a single char? Stuff like Killmails? Does SeAT pulls the Killmails also if the char can see the killmails?