Error: missing response (requestBody={"0":123,"1":34,"2":109,"3":101,"4":116,"5":104,"6":111,"7":100,"8"...

[sentry-io]

Sentry Issue: OUSD-4H

Error: missing response (requestBody={"0":123,"1":34,"2":109,"3":101,"4":116,"5":104,"6":111,"7":100,"8":34,"9":58,"10":34,"11":101,"12":116,"13":104,"14":95,"15":98,"16":108,"17":111,"18":99,"19":107,"20":78,"21":117,"22":109,"23":98,"24":101,"25":114,"26"...
  at Error: missing response (requestBody={"0":123,"1":34,"2":109,"3":101,"4":116,"5":104,"6":111,"7":100,"8":34,"9":58,"10":34,"11":101,"12":116,"13":104,"14":95,"15":98,"16":108,"17":111,"18":99,"19":107,"20":78,"21":117,"22":109,"23":98,"24":101,"25":114,... (/v2/[Key_removed_by_sparrowDom]%22,%20code=SERVER_ERROR,%20version=web/5.0.3))
  at t.makeError (/_next/static/chunks/0d2a819d.b3115171f16a910a5283.js:1:46000)
  at t.throwError (/_next/static/chunks/0d2a819d.b3115171f16a910a5283.js:1:46142)
  at None (/_next/static/chunks/0d2a819d.b3115171f16a910a5283.js:1:352143)
  at None (/_next/static/chunks/0d2a819d.b3115171f16a910a5283.js:1:349538)
...
(2 additional frame(s) were not displayed)

[micahalcorn]

Events: 150k
Users: 60

[franckc]

This is still occurring at a high frequency.
Does not seem to cause any breakage for the end users (we would have heard some complaints by now), so not a top priority.

CC @sparrowDom in case he knows what could be causing this...

[sparrowDom]

This looks to me as a service we use to query eth nodes. (wont say the name to not make it's keys extra easy to find by search. But the service shares a name with medieval chemistry like movement where the goal was to produce gold out of other substances and metals)

Anyway the prod access key was in the error report (have edited the comment and removed the key). Not sure we need to take any action here.

[franckc]

Interesting... I checked the service's dashboard and I don't see much failures. So I wonder if it's a timeout on the OUSd DApp side? e.g. we don't get a response and the DApp code times out?

[micahalcorn]

I'm seeing a ton of this along with the following errors on the client side:

[micahalcorn]

Events 424k
Users 92

I was unable to stake because of this issue (I think). After several page refreshes, the options eventually loaded.

[sparrowDom]

It is great that you caught this error on your machine @micahalcorn. As assumed seems like a problem with Alchemy. Will investigate if there is something we can do.

[sparrowDom]

I was unable to reproduce the issue locally. Some users have been having some luck (one got exact same error as we did) when hammering Infura / Alchemy with a lot of requests.

With this PR bumping the version of ethers to 5.0.30, that has a lot better error reporting when requests from providers fail.

[sparrowDom]

After getting some better error descriptions it seems this error boils down to:

• missing response error where Alchemy seems to block / no longer return data. Happens on multiple devices / environments with commonality that initial Alchemy requests are successful, later ones aren't. 24 users affected. I suspect a fix for this would be to reduce the amount of Alchemy requests made and Alchemy probably wouldn't be so trigger happy to block users.
• then there is the typeError error. Not sure what causes this error. Though weirdly Sentry seems to overreport on the number of events. See how these 3 events all have the same console output: event1, event2, event3. What is fishy is that all ~200 users affected by this in the last 14 days have the same Mobile Chrome version 76.0.3809, using Android 8.0.0. Which smells 🐟 to me (probably some kind of bot). For that reason I don't think it is worth addressing this problem.

Made a PR to further reduce Alchemy calls by doing a single API call when fetching all stablecoin balances (using Alchemy's special API) #562 .

[sparrowDom]

The one error we care about has almost gone away. 2-4 occurrences per day. The other I think we can ignore, since it contains the same Chrome version and android version and is most likely a bot.