Fork me on GitHub
#leiningen
<
2019-12-04
>
pdmct03:12:16

hi, I am using codebuild in AWS to build my app using Lein but for some of the dependencies I am getting connection resets on the pull from clojars. I have found some Maven issues that seem remarkably similar, their diagnosis was flaky container networking -- just wondering whether any one has seen this issue and whether there is simple fix to get it to do a few more retries

pdmct08:12:18

update: if I add a lein deps into my build script prior to the lein cljsbuild once I get this error in the deps call and it gets a second crack at clojars in the lein cljsbuild step and works ... huzzar! is there some way to work out what is happening here? or to make lein a bit more tolerant of the networking issues whatever they are?

regen11:12:56

I've had the same problem building my app in Travis. I ended up doing travis_retry lein deps, which makes Travis retry the command up to 5 times if it returns an error. I would also like to know if there's a way to make lein more lenient here!

mikerod15:12:34

Hmm I haven’t heard of this one before

mikerod15:12:13

Wonder if it’s been an issue reported in lein before - but sounds like it may just be related to the underlying aether lib used, or the “clj wrapping” layer pomegranate that lein uses

mikerod15:12:14

don’t see anything, but would think it’d relate to https://github.com/clj-commons/pomegranate/issues