mirrors/git-annex
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions

forwarded

Browse source
This commit is contained in:
Joey Hess 2017-08-15 13:53:51 -04:00
parent 35fc273218
commit e5109468e2
No known key found for this signature in database
GPG key ID: DB12DB0FF05F8F38
1 changed files with 29 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

29
doc/bugs/could_webdav_be_more_resilient_to_timeouts__63__/comment_1_bb4130ae787e19757b849a6301129c34._comment Normal file
View file

@ -0,0 +1,29 @@
[[!comment format=mdwn
username="joey"
subject="""comment 1"""
date="2017-08-15T17:02:20Z"
content="""
This last came up in 2014, at the time I disabled the http response timeout
entirely with setResponseTimeout Nothing.
Looking at Network.Protocol.HTTP.DAV.setResponseTimeout:
setResponseTimeout :: MonadIO m => Maybe Int -> DAVT m ()
#if MIN_VERSION_http_client(0,5,0)
setResponseTimeout rt = baseRequest %= \x -> x { responseTimeout = maybe responseTimeoutDefault responseTimeoutMicro rt }
#else
setResponseTimeout rt = baseRequest %= \x -> x { responseTimeout = rt }
#endif
Looks like with recent http-client versions in the case of Nothing, rather
than disabling the timeout, it now defaults to responseTimeoutDefault.
Which is 30 seconds. Since I'm passing it Nothing to try to disable the
timeout, that's kind of a problem!
I feel this is a bug in the DAV library. I could try to work around it
with `Just maxBound`, that's many years worth of microseconds on 64 bit, but
on 32 bit, it's somewhere under 10 minutes, which is really not good
enough when the goal is to disable timeouts at this level entirely.
Bug filed on DAV: <https://bugs.debian.org/872278>
"""]]