Date: Wed, 19 Sep 2018 12:14:49 +0000 From: Lorenzo Salvadore <phascolarctos@protonmail.ch> To: FreeBSD <freebsd-questions@freebsd.org> Subject: Re: error message with poudriere Message-ID: <AYVHRHPsISiOEIgpj7Avc0N9axdAQ3YeUhFC3ybFnOVFrKBXoVsDWJNQO8v-QdggxYksIrRFLgPdJFkuEP5fH-IEJpjUgwDfB3ei37WGgSQ=@protonmail.ch> In-Reply-To: <20180919074745.000015d6@aim.com> References: <20180919054924.00006e7f@aim.com> <H0ycS5ZvnYkv7HIfn_5ZRp_-JzwhmypKT0DavDZKHu2CDFmy0Cg18gCWBAi8b8v1q2k2ZbKxNlkS8WKFcaCHI6Vh4zyaGYSJ1WTzK2BUJf8=@protonmail.ch> <20180919074745.000015d6@aim.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> > > Running the devel/poudriere version of poudriere on a freebsd 11.2 am= d64 > > > machine, I am often presented with this message: > > > +swap_pager_getswapspace(18): failed > > > The number in parentheses changes, but the message stays the same. Th= e > > > message is printed literally hundreds of times. > > > Exactly what does it mean, and what do I do to correct it? I cannot f= ind > > > anything in the poudriere documentation that mentions it. > > > > I think the error means that you used all of you RAM and your swap spac= e. > > If I am right, then you need to add either more RAM or more swap space > > (easiest would probably be to create a swap file), unless there is a bu= g in > > ports-mgmt/poudriere-devel (for example a memory allocation in an infin= ite > > loop). > > I am sure you will soon get help from persons much more capable than me= , > > but while you wait I suggest you tell how much RAM and swap space do yo= u > > have, what are you trying to do with poudriere, if the problem arises w= ith > > ports-mgmt/poudriere too and send the output of the following two comma= nds: > > > > - swapinfo > > - sysctl vm.vmtotal > > > > The nearer you execute them to the the moment when the error is produce= d, > > the better it is. > > Lorenzo Salvadore. > > After a fresh reboot, this is the output: > > ~ # swapinfo > Device 1K-blocks Used Avail Capacity > /dev/ada0p2 2097152 0 2097152 0% > ~ # > ~ # sysctl vm.vmtotal > vm.vmtotal: > System wide totals computed every five seconds: (values in kilobytes) > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > > Processes: (RUNQ: 1 Disk Wait: 0 Page Wait: 0 Sleep: 119) > Virtual Memory: (Total: 4542784K Active: 4498692K) > Real Memory: (Total: 1662988K Active: 1656464K) > Shared Virtual Memory: (Total: 130504K Active: 96116K) > Shared Real Memory: (Total: 44132K Active: 37808K) > Free Memory: 794124K > > I was reading about a "PQ_LAUNDRY" variable/setting that pertains to Free= BSD > 11.x that is supposed to have some affect on this. I cannot find out much > information about it though or how to set it. I do not know about PQ_LAUNDRY (as I said, there is people much more skille= d than me), however I would suggest you to first exclude simpler solutions, if you have= not done it yet. The simplest solution is to consider the possibilty that your 4GB of RAM an= d 2GB of swap (if I have interpreted the data correctly) get full during your use of poud= riere: does that looks likely to you? If this is indeed the problem, then you should increase swap space. You can= do it by following these instructions: https://www.freebsd.org/doc/handbook/adding-swap-space.html If you are unsure that this is the problem, it might be better to first tes= t the solution by choosing to add swap through a swapfile (it is in the above documentatio= n link). Once tested, if this fixes your problem, then you might decide to use an ha= rd drive or a partion instead. Be prepared to have your system running slow: swap gives you more memory, b= ut reduces your computations speed. Lorenzo Salvadore.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AYVHRHPsISiOEIgpj7Avc0N9axdAQ3YeUhFC3ybFnOVFrKBXoVsDWJNQO8v-QdggxYksIrRFLgPdJFkuEP5fH-IEJpjUgwDfB3ei37WGgSQ=>