me
/
guix
Archived
1
0
Fork 0
Code

services: web: Pass run-directory to nginx's actions. 

Reloading nginx lead to (harmless, yet annoying) alert being logged:

    nginx: [alert] could not open error log file: open() "/gnu/store/9l2lznlqpjcm79znq1xas378maqgysc8-nginx-1.27.0/logs/error.log" failed (2: No such file or directory)

Since there already is prepared runtime directory for nginx, all that was
missing was to also pass it to the reload and reopen invocations.

* gnu/services/web.scm (nginx-shepherd-service)[reload, reopen]: Pass -p flag
with run-directory as value.

Change-Id: Id5c558d65def8c1fe5cf581a4a370508ac05e550
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
Tomas Volf 2024-07-05 17:01:20 +02:00 committed by Ludovic Courtès
parent 9afa6209ea
commit 01183a77aa
No known key found for this signature in database
GPG Key ID: 090B11993D9AEBB5
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
gnu/services/web.scm
View File

@ -857,11 +857,11 @@ This has the effect of killing old worker processes and starting new ones, using
the same configuration file. It is useful for situations where the same nginx
configuration file can point to different things after a reload, such as
renewed TLS certificates, or @code{include}d files.")
(procedure (nginx-action "-s" "reload")))
(procedure (nginx-action "-p" run-directory "-s" "reload")))
(shepherd-action
(name 'reopen)
(documentation "Re-open log files.")
(procedure (nginx-action "-s" "reopen"))))))))))
(procedure (nginx-action "-p" run-directory "-s" "reopen"))))))))))
(define nginx-service-type
(service-type (name 'nginx)