This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2022-11-25
Channels
- # announcements (8)
- # babashka (58)
- # beginners (59)
- # biff (4)
- # calva (39)
- # cider (2)
- # clj-kondo (8)
- # clj-together (4)
- # cljdoc (5)
- # cljsrn (1)
- # clojure (60)
- # clojure-australia (2)
- # clojure-europe (16)
- # clojure-nl (1)
- # clojure-norway (3)
- # clojurescript (13)
- # conjure (10)
- # cursive (9)
- # datomic (5)
- # dev-tooling (1)
- # emacs (6)
- # events (1)
- # graalvm (38)
- # graphql (5)
- # joyride (1)
- # kaocha (3)
- # lsp (23)
- # malli (2)
- # mount (2)
- # off-topic (31)
- # other-languages (13)
- # pathom (3)
- # polylith (12)
- # portal (4)
- # practicalli (22)
- # re-frame (6)
- # reagent (3)
- # releases (3)
- # sql (4)
- # squint (3)
- # tools-build (10)
- # tools-deps (10)
- # xtdb (4)
What’s the best way to go about debugging this issue?
datomic.client.api/transact api.clj: 200
datomic.client.api.protocols/fn/G protocols.clj: 72
datomic.client.api.sync/eval2257/fn sync.clj: 104
datomic.client.api.async/ares async.clj: 58
clojure.lang.ExceptionInfo: HttpConnectionOverHTTP@2580d230::DecryptedEndPoint@642196d5{l=/XXX.XXX.XXX.XXX:XXX,r=}
cognitect.anomalies/category: :cognitect.anomalies/unavailable
cognitect.anomalies/message: "HttpConnectionOverHTTP@2580d230::DecryptedEndPoint@642196d5{l=/XXX.XXX.XXX.XXX:XXX,r=}"
• It’s an exception that is thrown periodically. Maybe once every few hours or so.
• We have “Basic” monitoring enabled
• Running these stacks: https://s3.amazonaws.com/datomic-cloud-1/cft/973-9132/datomic-storage-973-9132.json & https://s3.amazonaws.com/datomic-cloud-1/cft/973-9132/datomic-compute-973-9132.json
• Nothing in the dashboard seems to look out of place or correlate with when these exceptions happen.
• No alerts in the logs or anything else that looks out of place.
• No cloudwatch alarms
• It’s happening in our production and staging environments which are completely separate deployments on different AWS accounts.
• Maybe it’s just a hiccup in the network however it seems too frequent.
I’m considering persisting the exception in-memory and then interrogating it through a remote REPL.
There is no monitoring enabled for API-Gateway. Maybe I could enable that?
Any other ideas?• Is this happening while you're transacting? Querying? Both? • How long is this request taking before it returns this exception? • What size instance are you using? • Do you have query groups or are all of these requests going to your primary compute group?
• Is this happening while you’re transacting? Querying? Both? ◦ Transacting • How long is this request taking before it returns this exception? ◦ Hard to tell, although it appears to be not long at all. < 100ms • What size instance are you using? ◦ t3.small - Dashboard shows CPUUtilization around 2% and memory below 800MB • Do you have query groups or are all of these requests going to your primary compute group? ◦ We have a compute stack and a storage stack only. So no query groups I believe
@U3XCG2GBZ Why don't you open a https://www.datomic.com/support.html with us and we can check it out after the break?
No worries