Re: go-1.16.2 out of memory when building Go written program
I've set ulimit -d 400 All builds have been done fine once changed. Martin Sent with ProtonMail Secure Email. ‐‐‐ Original Message ‐‐‐ On Wednesday, June 23, 2021 6:15 PM, Sven F. wrote: > On Wed, Jun 23, 2021 at 2:03 PM Martin martin...@protonmail.com wrote: > > > Hi list, > > I try to build terraform-provider-aws and terraform-provider-google. > > $ go build > > produces an error "out of memory" . > > May it be malloc related issue or how to fix it in other way? > > Thank you for answer in advance. > > Martin > > man login.conf > > - > > -- > > --- > > Knowing is not enough; we must apply. Willing is not enough; we must do
Re: mime type not set correctly for webpage
On Mon, Jun 21, 2021 at 5:35 AM Stuart Henderson wrote:
>
SNIP
>
> The location matches this block:
>
> >> location "/rainloop/*" {
> >> fastcgi param SCRIPT_FILENAME "/htdocs/rainloop/index.php"
> >> fastcgi socket "/run/php-fpm.sock"
> >> # directory index "index.php"
> >> }
>
> so it ought to be served by rainloop's front controller in
> /htdocs/rainloop/index.php and thus the content-type ought to be
> dictated by rainloop..
>
> I am wondering about the doubled path entry /rainloop/rainloop/ and
> whether that is involved in the problem, perhaps making it serve
> an error page not the expected css. (Opening that url directly
> would prove that).
>
> If so, perhaps httpd.conf "strip" might help (though probably the
> bit that wants a component stripping is actually PATH_INFO and the
> documentation doesn't talk about changing that, but it maybe the
> doc rather than the code that is incomplete).
I thought the config was a little odd, but that was the config based
upon rainloop info.
If I bypass index.php and access css file with direct URL the file is
found, but it still has the incorrect mime type associated with it.
thanks
diana
Re: mime type not set correctly for webpage
inline
On Sun, Jun 20, 2021 at 8:55 AM Joseph Olatt wrote:
>
> On Thu, Jun 17, 2021 at 10:19:12PM -0600, Diana Eichert wrote:
> > "I'm trying to get rainloop PHP webmail setup on a mail server using
> > OpenBSD 6.9 httpd
> >
> > I have the webserver configured however the browser shows the mime
> > type is not correct for the css style sheet.
> > >From Firefox console,
> > "The stylesheet
> > https://mail.shopnet.com/rainloop/rainloop/v/1.16.0/static/css/app.min.css?standard
> > was not loaded because its MIME type, ???text/html???, is not
> > ???text/css???.
> >
> > I included system mime file types in httpd.conf file as below.
> >
> > Any pointers are appreciated.
>
> My suspicion is that the extension is being assumed to be
> ".css?standard".
>
> So, try changing the following line in /usr/share/misc/mime.types:
>
> text/csscss
>
> to:
>
> text/csscss css?standard
>
> restart httpd and see if resolves your problem.
>
> This is just a thought.
I tried that, had the same issue.
thanks
> >
> > thanks
> >
> >
> > # $OpenBSD: httpd.conf,v 1.20 2018/06/13 15:08:24 reyk Exp $
> >
> > types { include "/usr/share/misc/mime.types" }
> >
> > server "mail.shopnet.com" {
> > listen on * port 80
> > location "/.well-known/acme-challenge/*" {
> > root "/acme"
> > request strip 2
> > }
> > location * {
> > block return 302 "https://$HTTP_HOST$REQUEST_URI;
> > }
> > }
> >
> > server "mail.shopnet.com" {
> > listen on * tls port 443
> > tls {
> > certificate "/etc/ssl/mail.shopnet.com.fullchain.pem"
> > key "/etc/ssl/private/mail.shopnet.com.key"
> > }
> > location "/pub/*" {
> > directory auto index
> > }
> > location "/.well-known/acme-challenge/*" {
> > root "/acme"
> > request strip 2
> > }
> > location "/rainloop/*" {
> > fastcgi param SCRIPT_FILENAME "/htdocs/rainloop/index.php"
> > fastcgi socket "/run/php-fpm.sock"
> > # directory index "index.php"
> > }
> >
> > # Value below is 25MB in bytes. 1MB = 1048576 bytes
> > connection max request body 26214400
> > location "/data*" {
> > block return 403
> > }
> >
> > location "/rainloop/*.php*" {
> > fastcgi socket "/run/php-fpm.sock"
> > }
> > }
go-1.16.2 out of memory when building Go written program
Hi list, I try to build terraform-provider-aws and terraform-provider-google. $ go build produces an error "out of memory" . May it be malloc related issue or how to fix it in other way? Thank you for answer in advance. Martin
support and consulting: new entry request
0 C Japan P Osaka T Osaka Z O Scqr Inc. I A M contact@obsd.solutions U https://obsd.solutions/ B X N We are experienced ICT designers/developers and security monks, using OpenBSD as primary servers for years. [ Company Website ] https://www.scqr.net/;>scqr.net
Re: go-1.16.2 out of memory when building Go written program
On Wed, Jun 23, 2021 at 2:03 PM Martin wrote: > > Hi list, > > I try to build terraform-provider-aws and terraform-provider-google. > > $ go build > > produces an error "out of memory" . > > May it be malloc related issue or how to fix it in other way? > > Thank you for answer in advance. > > Martin > man login.conf -- -- - Knowing is not enough; we must apply. Willing is not enough; we must do
Re: altroot weekly.local
Stuart Henderson writes: > I would prefer to use almost anything else though and get versioned > backups. Probably my most used backup/restore action is to get back a > version of some file from yesterday so something that will only write > the changes is useful. I quite like borg for this but there are many > options. I use a slightly modified version of the script here to have a 7-day versioned backup of important files. https://lika.be/wp/2011/06/incremental-backup-script-with-daily-delta/ You can find it in various other sites, original credit seems to be tri...@linuxcare.com I've tried duplicity and one or two other more sophisticated backup tools. I like this because it uses only one very standard utility, rsync, and I find it easy to reason about and (importantly) easy to restore from. Allan
old zlib version and deflateBound function
Hi, actually, when using fwrite function in R package data.table, an error occurs because the value returned by deflateBound function from libz (part of the system) is incorrect. This point seems fixed in zlib version 1.2.3.1 "Take into account wrapper variations in deflateBound()". DeflateBound() in current libz version (1.2.11) has been tested on many platforms and is OK. More details can be found on issue "fwrite gzip on OpenBSD error" on GitHub (https://github.com/Rdatatable/data.table/issues/5048) R data.table team accepted a patch that resolved the problem : openBSD succeeds now in all tests (with no OpenMP support reported when installing package). With this message, they can place a link in news item so that readers of data.table news can be reassured that it has been reported to OpenBSD team. Thanks, --- PhC
Re: rpki-client and BLACKHOLE routes
On Wed, Jun 23, 2021 at 11:40:25AM +0200, Hrvoje Popovski wrote: > Hi all, > > fist of all, thank you for rpki-client, it's so easy to use it and to > get the job done. > I'm playing with rpki-client and denying ovs invalid statement and I've > seen that with default ovs config statement (deny from ebgp ovs invalid) > BLACKHOLE routes are blocked/invalid. > > What is the right way to allow BLACKHOLE routes through rpki ? Or if > someone can give me a hint on what to do. > BLACKHOLE routes normally have a more specific check so you can re-allow them back after the ovs invalid check (for that you need to take away the quick from the default ruleset or actually allow quick the blackholes before). I guess you can use something along the lines of: allow quick from group clients inet prefixlen 32 community $BLACKHOLE set nexthop blackhole allow quick from group clients inet6 prefixlen 128 community $BLACKHOLE set nexthop blackhole I guess you also have some client prefix-sets that should be added to the filter rule so that one client can not blackhole for another. BLACKHOLE routes are done in many ways and I'm not sure if there is consensus who is allowed to announce what. Also if there are multiple paths to the destination should the blackhole only be active if the covering route is from the same peer? -- :wq Claudio
rpki-client and BLACKHOLE routes
Hi all, fist of all, thank you for rpki-client, it's so easy to use it and to get the job done. I'm playing with rpki-client and denying ovs invalid statement and I've seen that with default ovs config statement (deny from ebgp ovs invalid) BLACKHOLE routes are blocked/invalid. What is the right way to allow BLACKHOLE routes through rpki ? Or if someone can give me a hint on what to do. Thank you...
Re: altroot weekly.local
On 2021-06-22, Andrew Robertson wrote: > Is there any problem with putting ROOTBACKUP=1 in my weekly.local > instead of daily.local? I'm backing up to an SD card and it's maybe not > fast enough to back up in 24 hours, plus weekly backup would be fine. It won't do anything in weekly.local; the weekly script doesn't handle ROOTBACKUP. You could use a shell conditional in daily.local instead: [ $(date +%a) == Sun ] && ROOTBACKUP=1 I advise against using ROOTBACKUP here though. It just dd's the filesystem. If the run is taking a long time, there's a high chance that your backup will not be consistent, both due to the filesystem changing during a backup, and due to overwriting the old backup with the new one. I would prefer to use almost anything else though and get versioned backups. Probably my most used backup/restore action is to get back a version of some file from yesterday so something that will only write the changes is useful. I quite like borg for this but there are many options.
