| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
eof while reading, errno 0
using check_curl on a probably embedded device responding as 'Server: GoAhead-Webs'
%> check_curl -H ... -S -vvv
> GET / HTTP/1.1
Host: ...
User-Agent: check_curl/v2.4.0 (monitoring-plugins 2.4.0, libcurl/7.76.1 OpenSSL/3.0.7 zlib/1.2.11 brotli/1.0.9 libidn2/2.3.0 libpsl/0.21.1 (+libidn2/2.3.0) libssh/0.10.4/openssl/zlib nghttp2/1.43.0)
Accept: */*
Connection: close
* Mark bundle as not supporting multiuse
* HTTP 1.0, assume close after body
< HTTP/1.0 302 Redirect
< Server: GoAhead-Webs
< Date: Tue Mar 26 17:57:16 2019
< Cache-Control: no-cache, no-store, must-revalidate,private
< Pragma: no-cache
< Expires: 0
< Content-Type: text/html
< X-Frame-Options: sameorigin
< X-XSS-Protection: 1; mode=block
< X-Content-Type-Options: nosniff
< Location: https://...
<
* OpenSSL SSL_read: error:0A000126:SSL routines::unexpected eof while reading, errno 0
* Closing connection 0
reading the discussion on https://github.com/openssl/openssl/discussions/22690 suggest to set the option SSL_OP_IGNORE_UNEXPECTED_EOF
which makes check_curl behave like check_http at this point.
Since this is a rather new flag, fencing it in ifdefs.
And since there can only be one ssl ctx function, we need to move both tasks into one function.
|
|
|
|
|
|
|
|
Previously the --state-regex option accepted only "critical" and
"warning" as values.
This commit changes the strcmp there to strcasecmp to be more tolerant
regarding the input.
|
|
The help output of `check-curl` contained a typo,
the real option is `state-regex` and not `regex-state` as
the help suggests.
Also added the two possible options to avoid confusion.
|
|
From the mere help output for -C / --certificate, I was confused about
what its two integer parameters do. Unfortunately, I also missed out on
the explaining examples later. Since I like to have basic documentation
for each flag, I tried to make the arguments as short as possible.
The other fix was one hyphen too many for the --cookie-jar option.
|
|
This change aims to raise the worst status between the SSL check and the HTTP check.
before:
check_curl -H www.google.fr -S --continue-after-certificate --certificate 4000,4000 ; echo $?
CRITICAL - Certificate '*.google.fr' expires in 74 day(s) (Tue 22 Oct 2024 12:53:52 PM GMT +0000).
HTTP OK: HTTP/2 200 - 22807 bytes in 0.076 second response time |time=0.075516s;;;0.000000;10.000000 size=22807B;;;0;
0
after:
/usr/lib/nagios/ovh/check_curl -H www.google.fr -S --continue-after-certificate --certificate 4000,4000 ; echo $?
CRITICAL - Certificate '*.google.fr' expires in 74 day(s) (Tue 22 Oct 2024 12:53:52 PM GMT +0000).
HTTP OK: HTTP/2 200 - 22840 bytes in 0.090 second response time |time=0.090463s;;;0.000000;10.000000 size=22840B;;;0;
2
|
|
Having a webserver respond with a relative redirect as for ex. in `Location: /path/to.html`
check_curl would use the wrong standard http/https port instead
of crafting the absolute url using the given scheme/hostname and port.
Adding a new test case for this for check_http and check_curl. check_http did
it correct already, so no fix necessary there.
before:
%>./check_curl -H 127.0.0.1 -p 50493 -f follow -u /redirect_rel -s redirected -vvv
**** HEADER ****
HTTP/1.1 302 Found
...
Location: /redirect2
...
* Seen redirect location /redirect2
** scheme: (null)
** host: (null)
** port: (null)
** path: /redirect2
Redirection to http://127.0.0.1:80/redirect2
fixed:
%>./check_curl -H 127.0.0.1 -p 50493 -f follow -u /redirect_rel -s redirected -vvv
**** HEADER ****
HTTP/1.1 302 Found
...
Location: /redirect2
...
* Seen redirect location /redirect2
** scheme: (null)
** host: (null)
** port: (null)
** path: /redirect2
Redirection to http://127.0.0.1:50493/redirect2
Signed-off-by: Sven Nierlein <sven@nierlein.de>
|
|
regex check
from the default CRITICAL to something else (#1213)
|
|
|
|
|
|
Fix compiler warning for uninitialized variable
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* lib/utils_base.c
* plugins/check_curl.c
* plugins-root/check_dhcp.c
Removed a line which theoretically can not do anything, but there was
comment which indicated something else. Still trying this though.
|
|
|
|
warning
|
|
|
|
|
|
|
|
first getaddrinfo result which succeeds
|
|
|
|
|
|
|
|
|
|
|
|
|
|
curl >= 7.85.0
|
|
|
|
check_curl fails on large pages:
HTTP CRITICAL - Invalid HTTP response received from host on port 5080: cURL returned 23 - Failure writing output to destination
for example trying to run check_curl on the test from #1822
I guess the idea is to double the buffer size each time it is to small. But the code
exponentially grows the buffer size which works well 2-3 times, but then fails.
|
|
* If server_address is an IPv6 address surround it with brackets
* If the message is too short, we should not have an underflow
* Add simple conditional test case available if IPv6 is
|
|
|
|
PicoHTTPParser
|
|
|
|
check_curl.c: bugfix: verify certificates option should not force SSL to be used
|
|
|
|
errors)
|
|
|
|
|
|
|
|
- added verbose output in verify_callback
- pin refcounting for certs (avoid subject extraction error when checking
certs in is_openssl_callback mode)
|
