Alexander A. Klimov
17ce004abd
Benchmark result set read
2022-03-29 11:16:59 +02:00
Alexander A. Klimov
e3a29fb072
Checkable#ExecuteEventHandler(): don't outsource event command run twice
...
refs #8704
2022-03-29 11:12:53 +02:00
Alexander A. Klimov
eddb4beea0
Log can't-do-yets
2022-03-28 17:11:46 +02:00
Alexander A. Klimov
10b3bacb48
Revert "Log can't-do-yets summmary"
...
This reverts commit a652c4f52a1ac713bcfcd87979d9adae595673f6.
2022-03-28 17:07:17 +02:00
Alexander A. Klimov
06e0d75f99
Benchmark FinishAsyncQueries
2022-03-28 17:06:43 +02:00
Alexander A. Klimov
a652c4f52a
Log can't-do-yets summmary
2022-03-28 15:33:27 +02:00
Alexander A. Klimov
4d90cc88af
Revert "Log can't-do-yets"
...
This reverts commit da197960d13bea11bc1d89043799c066ca470baf.
2022-03-28 15:20:46 +02:00
Alexander A. Klimov
5208511a6f
TxTimerHandler!
2022-03-28 13:47:15 +02:00
Alexander A. Klimov
8f58da8a0f
(A)sync? per query
2022-03-28 13:46:09 +02:00
Alexander A. Klimov
da197960d1
Log can't-do-yets
2022-03-28 13:40:43 +02:00
Alexander A. Klimov
9cda9396bf
Benchmark actual query exe.
2022-03-28 13:36:38 +02:00
Julian Brost
c2abb0c9dc
GelfWriter: show error message of exceptions
2021-08-17 18:49:36 +02:00
Julian Brost
8da90d44fa
InfluxdbWriter: actually verify TLS server certificates
...
And add a new option ssl_insecure_noverify to explicitly disable it if desired.
2021-08-17 18:49:36 +02:00
Julian Brost
037944a51b
GelfWriter: actually verify TLS server certificates
...
And add a new option insecure_noverify to explicitly disable it if desired.
2021-08-17 18:49:35 +02:00
Julian Brost
78aa348e6a
ElasticsearchWriter: actually verify TLS server certificates
...
And add a new option insecure_noverify to explicitly disable it if desired.
2021-08-17 18:49:35 +02:00
Julian Brost
5c35ab551b
Enable hostname verification in UnbufferedAsioTlsStream
2021-08-17 18:49:35 +02:00
Alexander A. Klimov
caf697b868
API: hide ApiListener#ticket_salt
2021-07-09 09:30:01 +02:00
Julian Brost
5f9957c7d5
Remove passwords from API
...
IdoMysqlConnection, IdoPgsqlConnection, and ElasticsearchWriter require
passwords in their configuration to authenticate against external services.
This commit ensures that these can no longer be accessed using the API.
2021-07-09 09:05:40 +02:00
Julian Brost
aac5946d5d
Merge pull request #8869 from Icinga/feature/improve-crashlog-211
...
Improve crashlog
2021-07-09 08:21:54 +02:00
Julian Brost
58be537557
Merge pull request #8891 from Icinga/bugfix/trigger-fixed-downtimes-immediately-211
...
Downtime#Start(): trigger fixed downtimes immediately instead of waiting for the timer
2021-07-08 15:33:52 +02:00
Julian Brost
a261a96a99
Add comments to stack trace formatter and test case
2021-07-08 13:59:39 +02:00
Julian Brost
3d650e865d
Use backtrace_symbols() when printing stack traces on FreeBSD
...
Unfortunately, the symbol resolution of boost::stacktrace is broken on
FreeBSD, therefore fall back to using backtrace_symbols() to print the
stack trace saved by Boost.
Additionally, -D_GNU_SOURCE is required on FreeBSD for the
_Unwind_Backtrace function used by boost::stacktrace.
2021-07-08 13:59:39 +02:00
Julian Brost
0246765a5c
Crash handlers: use more compact string representation
2021-07-08 13:59:39 +02:00
Julian Brost
0dda2de696
Begin crash log for SIGABRT with error message and timestamp
...
This makes the format more similar to what the uncaught C++ and SEH
exception handlers write. Previously there was no indication in the
crash log that a SIGABRT happened.
2021-07-08 13:59:39 +02:00
Julian Brost
a342d75cc3
Add some comments to __cxa_throw
...
Maybe this will save the next person who has to look at this code some
time. Please don't blame me for the implementation, I'm just trying to
reconstruct what it does.
2021-07-08 13:59:39 +02:00
Julian Brost
9aeb962863
Restructure stack and context trace selection in DiagnosticInformation and document behavior
...
The logic for selecting the traces to print stays the same, but there
are fewer nested ifs now. This changes the format of the returned string
a bit by adding a heading for both traces.
2021-07-08 13:59:39 +02:00
Julian Brost
6104df37dc
Add documentation for cast_exception function
2021-07-08 13:59:39 +02:00
Julian Brost
27767b1aa3
Pass fallback stacktrace to DiagnosticInformation in terminate handler
...
By default, DiagnosticInformation uses the stack trace saved when the
exception was thrown, but this mechanism is not in use on Windows.
Gathering a stacktrace in the terminate handler serves as a fallback.
2021-07-08 13:59:39 +02:00
Julian Brost
58d553a971
Replace icinga::StackTrace with boost::stacktrace::stacktrace
...
Provides roughly the same functionality but works better on certain
platforms (especially Windows) and is less code to maintain.
2021-07-08 13:59:39 +02:00
Julian Brost
fb303f5205
Print details in uncaught SEH exception handler
2021-07-08 13:59:39 +02:00
Julian Brost
a77c37da8e
Move error message and time to the beginning of the SEH crash log
...
This is more similar to the normal exception crashlog which also states
the problem and time at the beginning of the file.
2021-07-08 13:59:39 +02:00
Julian Brost
2c9b1d8415
Use boost::stacktrace instead of custom implementation in Windows SEH filter
2021-07-08 13:59:39 +02:00
Julian Brost
0b9ef5ab6d
Run termination handler for uncaught C++ exceptions on Windows
...
On Windows, the termination handler is executed for uncaught C++
exceptions unless a SEH unhandled exception filter is also set. In this
case, this filter has to explicitly chain the default filter to keep
this behavior.
2021-07-08 13:59:39 +02:00
Alexander Aleksandrovič Klimov
7c4996d1c2
Merge pull request #8878 from Icinga/bugfix/fixed-dt-end-notify-211
...
Checkable::NotifyDowntimeEnd(): don't send Downtime end notification unless triggered
2021-07-07 22:37:04 +02:00
Alexander A. Klimov
05b2606fbb
Checkable::NotifyDowntimeEnd(): don't send Downtime end notification unless triggered
...
... for fixed Downtimes as well.
2021-07-07 15:44:48 +02:00
Alexander A. Klimov
084acbe028
On ScheduledDowntime change: remove future downtimes created before change
...
refs #8309
2021-07-07 15:44:46 +02:00
Alexander A. Klimov
ffa3872b28
On ScheduledDowntime change: ignore downtimes created before change
...
... while creating new downtimes.
refs #8309
2021-07-07 15:44:46 +02:00
Alexander A. Klimov
a262d18921
Introduce Downtime#config_owner_hash
...
refs #8309
2021-07-07 15:44:46 +02:00
Alexander A. Klimov
743af40114
ScheduledDowntime: ignore not related Downtimes while creating Downtimes
2021-07-07 15:44:46 +02:00
Alexander A. Klimov
4216ed018c
Downtime#Start(): trigger fixed downtimes immediately instead of waiting for the timer
...
... not to cause e.g. notifications if a problem occurs
between the downtime start time and the timer routine.
2021-07-07 15:34:04 +02:00
Noah Hilverling
013ff97ae1
Merge pull request #8870 from Icinga/bugfix/downtime-for-host-service-with-long-name-8022-211
...
ConfigObjectUtility::GetObjectConfigPath(): hash names of not already existing objects
2021-07-07 10:39:03 +02:00
Noah Hilverling
13f091272f
Merge pull request #8871 from Icinga/bugfix/harden-scheduled-downtimes-211
...
ScheduledDowntime::TimerProc(): Catch exceptions to make sure other downtimes are still created
2021-07-07 10:38:57 +02:00
Noah Hilverling
f444d1e4f9
ScheduledDowntime::TimerProc(): Catch exceptions to make sure other downtimes are still created
2021-07-06 12:27:57 +02:00
Julian Brost
2131318c48
Add timeout for all new connections
...
This commit adds a timeout for both establishing new outgoing and incoming
connections. This timeout applies to everything until the connection is in a
state where either JsonRpcConnection or HttpServerConnection takes over.
2021-07-06 11:33:48 +02:00
Julian Brost
2dc5c9e47b
GetObjectConfigPath: only truncate and hash comment and downtime filenames
...
This partially reverts 68a0079c26686363b6202a8abd2712d2bf96d9f2 and keeps the
fix only for comment and downtime objects for now. For reasoning, please see
the comment in the code.
2021-07-05 16:18:40 +02:00
Julian Brost
3db48de0e6
GetObjectPath: ensure use of escaped name in all cases and use TruncateUsingHash()
...
68a0079c26686363b6202a8abd2712d2bf96d9f2 introduced two problems that are fixed
with this commit:
1. The new truncated/hashed name did not use EscapeName()
2. There was a possible collision of names when creating objects with a full
name of format "[80 characters]...[40 hex digits]" (i.e. the same as the
truncated/hashed variant but short enough that it isn't hashed)
2021-07-05 16:18:40 +02:00
Julian Brost
fc0019b271
Utility: add a function to truncate strings while avoiding collisions
2021-07-05 16:18:40 +02:00
Alexander A. Klimov
c174456ed3
ConfigObjectUtility::GetObjectConfigPath(): hash names of not already existing objects
...
... to avoid too long file names.
refs #8022
2021-07-05 16:18:40 +02:00
Alexander Aleksandrovič Klimov
0bf2c78cc4
Merge pull request #8810 from Icinga/bugfix/mysql-one-transaction-for-programstatus-2.11
...
IDO: Use own transaction for program status and make sure InternalNewTransaction() gets executed
2021-05-26 15:46:56 +02:00
Alexander Aleksandrovič Klimov
5f6fb3d617
Merge pull request #8770 from Icinga/bugfix/retry-rename-on-windows-2.11
...
Retry file rename operations on Windows for some errors (2.11)
2021-05-26 13:43:13 +02:00