[RESOLVED] Degraded ESI Performance

EVE Information Portal Announcements
1

We are currently investigating degraded ESI performance after this morning’s maintenance work. Keep an eye on status.eveonline.com for relevant updates. Thank you for your patience.

Updates

  • 14:00 UTC - The issue has been identified and performance is starting to recover. We will continue to keep a close eye on it.
  • 14:15 UTC - All services have restarted and performance has recovered. We will continue to monitor.
8 Likes
2

Ok, keep up the good work!

3 Likes
3

there come the killmails =^_^=

4

What makes a zkillmail worth looking at?

There are other reasons for ESI than just silly killmails

5

There is still issue with API? I get lot of [{“error”:“Timeout contacting tranquility”,“timeout”:10}].

6

yeah a lot of errors, even when the request actually succeeds eg for ui POST

1 Like
7

I must finally get around to better error handling :smiling_face_with_tear:

8

Always had to.

9

Many problems now getting

wallet : {“error”:“Timeout contacting tranquility”,“timeout”:10}

In this case is wallet, but can pass in many of them.

10

@CCP_Lumi esi issues returned.

edit: seems resolved again

11

I have again problems with ESI. Timeout in some queries ( i think assets but dont pintpointing now because not in my developer pc )

12

The problem still exists.
Here is an example where I get “Timeout contacting tranquility” on the ESI interface description itself:

https://esi.evetech.net/ui/#/Market/get_markets_region_id_orders

test
test1438×1105 86.9 KB

On the second attempt it worked for the same typeid but the error message appears roughly 1-2 times after 10 attempts.

I hope the problem will be solved because the only alternative for me would be to make this call without TypeID (and thus Jita complete) and that would be 380 pages each time, which in turn increases the traffic enormously.

Regards

13

That’s actually what you should do.

14

Can I call this interface every 300 seconds with a clear conscience?
(I would have thought it would be smarter if I only determined the one required TypeID and not everything consecutively.)

15

I make a pinpointing, currently the mails esi endpoint is failing. I use for the recruiting bit of the operation.

16

You should.
Remember that the market are cached on the server side, so your 300 pages will hit the cache and return an already present page.

Meanwhile, if you call each type page, it’s more likely to be a miss and request the server access.