[Nix-dev] hydra admins: please take care of i686 GHC out of memory issues

[Vladimír Čunát]

Hi.

To me it's clear that some other job also depends on it (and got built 
first), so that's why it's "cached".

On 01/22/2014 11:44 PM, Gergely Risko wrote:
> In the meantime can we please disable any parallelism during build for
> i686 (so no -j passing to any stage of the compiler) and try like that?
> Maybe we're just somehow hitting the 4G one process address space limit
> with threads or something.

Parallel build already *was* disabled for GHC.
Cf. rev. 28483364 for the default 7.6.3 and rev. cb2168 for most others.

> Also, would it too much of a waste of hydra resources to try 10
> consecutive builds for_absolutely the same derivation_  and gather
> failure/success ratios?

I think there was some discussion about auto-restarting builds with 
transient failures, but I don't remember what's the status/conclusions...


Vlada


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3251 bytes
Desc: S/MIME Cryptographic Signature
Url : http://lists.science.uu.nl/pipermail/nix-dev/attachments/20140123/6470884a/attachment.bin