Commit Graph

6448 Commits

Author SHA1 Message Date
Alexander A. Klimov c0f87dd4c9 /v1/actions/schedule-downtime: reject request on invalid trigger_name
For this purpose lookup the specified Downtime. Also pass Downtime objects,
not just names, to Downtime::AddDowntime() not to lookup it twice.
2024-04-25 12:13:52 +02:00
Alexander A. Klimov f0b5239a15 [Refactor] Downtime::GetDowntimeIDFromLegacyID(): return the Downtime itself
not just its name.
2024-04-25 12:13:52 +02:00
Alexander A. Klimov 28b0f7a48c [Refactor] l_LegacyDowntimesCache: store Downtime objects, not just their names
to avoid names of vanished objects.
2024-04-24 12:33:56 +02:00
Alexander A. Klimov bb13e98ca5 PluginCheckTask::ProcessFinishedHandler(): warn about exit codes outside 0..3
in the plugin output as well, in addition to the warning log.
2024-04-23 17:45:31 +02:00
Alexander A. Klimov e33befabfb Make ProcessResult#ExitStatus and CheckResult#exit_status 64-bit ints
so that they can hold Windows exit codes like 3221225477 (>2147483647).
2024-04-23 17:45:31 +02:00
Alexander A. Klimov 5c17465a19 OpenTsdbWriter#CheckResultHandler(): skip custom tags with empty values
refs #7724
2024-04-18 11:36:21 +02:00
Julian Brost 31be43ff6c
Merge pull request #10018 from Icinga/revert-9980-config-sync-conflicts
Revert "Process `config::update/delete` cluster events gracefully"
2024-03-08 16:58:28 +01:00
Julian Brost af97431bfb
Merge pull request #10006 from Icinga/http-error-handling
HttpServerConnection: use exceptions for error handling
2024-03-08 15:06:51 +01:00
Yonas Habteab a924a49cd8
Revert "Process `config::update/delete` cluster events gracefully" 2024-03-07 17:17:17 +01:00
Julian Brost 097ba00a9c
Merge pull request #10008 from Icinga/Al2Klimov-patch-12
Don't unnecessarily shuffle items before config validation
2024-03-07 16:44:38 +01:00
Alexander Aleksandrovič Klimov 629038344b
OpenTsdbWriter#CheckResultHandler(): clarify log messages
Clarify which "host or service" an "Unable to resolve macro" debug log message refers to.
2024-02-22 10:34:35 +01:00
Julian Brost abea2f270c
Merge pull request #9997 from Icinga/ListenerCoroutineProc-remote_endpoint
ApiListener#ListenerCoroutineProc(): get remote endpoint ASAP for logging
2024-02-20 13:46:02 +01:00
Alexander Aleksandrovič Klimov 51cdd593da
Don't unnecessarily shuffle items before config validation
Before ae693cb7e1 (#9577) we've repeatedly looped over all items in parallel like this:

while not types.done:
  for t in types:
    if not t.done and t.dependencies.done:
      with parallel(all_items, CONCURRENCY) as some_items:
        for i in some_items:
          if i.type is t:
            i.commit()

I.e. all items got distributed over CONCURRENCY threads, but not always equally. E.g. it was the hosts' turn, but only two threads got hosts and did all the work. The others didn't do actual work (due to the lack of hosts in their queue) which reduced the performance. c721c302cd (#6581) fixed it by shuffling all_items first. ae693cb7e1 (#9577) made the latter unnecessary by replacing the above algorithm with this:

while not types.done:
  for t in types:
    if not t.done and t.dependencies.done:
      with parallel(all_items[t], CONCURRENCY) as some_items:
        for i in some_items:
          if i.type is t:
            i.commit()

I.e. parallel() gets only items of type t, so all threads get e.g. hosts.
2024-02-19 14:26:06 +01:00
Julian Brost 700c5a13d7 HttpServerConnection: use exceptions for error handling
When a HTTP connection dies prematurely while the response is sent,
`http::async_write()` sets the error code to something like broken pipe for
example. When calling `async_flush()` afterwards, it sometimes happens that
this never returns. This results in a resource leak as the coroutine isn't
cleaned up. This commit makes the individual functions throw exceptions instead
of silently ignoring the errors, resulting in the function terminating early
and also resulting in an error being logged as well.
2024-02-19 14:12:41 +01:00
Julian Brost 04ef105caa
Merge pull request #9980 from Icinga/config-sync-conflicts
Process `config::update/delete` cluster events gracefully
2024-02-19 13:49:41 +01:00
Julian Brost 7d1c887a32
Merge pull request #9999 from Icinga/reset-log-message-count-correctly
ApiListener: Reset `m_LogMessageCount` when rotating
2024-02-15 17:06:16 +01:00
Alexander Aleksandrovič Klimov 9db1c4aca3
Merge pull request #8011 from Icinga/bugfix/reset-sigpipe-6912
Reset all signal handlers of child processes
2024-02-15 12:22:36 +01:00
Yonas Habteab 456144c1dc ApiListener: Process cluster config updates sequentially 2024-02-14 14:25:53 +01:00
Yonas Habteab 40011b0584 Introduce `ObjectNamesMutex` helper class 2024-02-14 14:25:53 +01:00
Alexander Aleksandrovič Klimov 1a8ce5a90e
Merge pull request #9575 from Icinga/WorkQueue-ParallelFor
WorkQueue#ParallelFor(): allocate lambda once per thread, not once per item
2024-02-14 12:59:50 +01:00
Julian Brost 2be08aa2e0
Merge pull request #9992 from Icinga/remove-redundat-cpu-bound-work
Drop redundant `CpuBoundWork` usage in `JsonRpcConnection::Disconnect()`
2024-02-13 15:51:34 +01:00
Julian Brost fc6a106345
Merge pull request #9994 from Icinga/redundant-cpu-bound-work-usages
Drop redundant `CpuBoundWork` usages in `lib/remote`
2024-02-13 14:53:59 +01:00
Alexander Aleksandrovič Klimov 48eb563ca0
Merge pull request #9736 from Icinga/stream-read-allow_partial
Stream#Read(): remove de facto unused param allow_partial
2024-02-13 13:04:15 +01:00
Yonas Habteab 008fcd1744 Preserve runtime objects in a tmp file for the entire validation process
Given that the internal `config::Update` cluster events are using this
as well to create received runtime objects, we don't want to persist
first the conf file and the load and validate it with `CompileFile`.
Otherwise, we are forced to remove the newly created file whenever we
can't validate, commit or activate it. This also would also have the
downside that two cluster events for the same object arriving at the
same moment from two different endpoints would result in two different
threads simultaneously creating and loading the same config file -
whereby only one of the surpasses the validation, while the other is
facing an object `re-definition` error and tries to remove that config
file it mistakenly thinks it has created. As a consequence, an object
successfully created by the former is implicitly deleted by the latter
thread, causing the objects to mysteriously disappear.
2024-02-12 15:18:32 +01:00
Yonas Habteab 6e66cd9aff ApiListener: Reset `m_LogMessageCount` when rotating
Closing and re-opening that very same log file shouldn't reset the
counter, otherwise some log files may exceed the max limit per file as
their offset indicator is reset each time they are re-opened.
2024-02-09 18:04:20 +01:00
Yonas Habteab eb813cfb99 HttpServerConnection: Drop superfluous `CpuBoundWork` usage 2024-02-09 15:17:26 +01:00
Alexander A. Klimov 62e1d7650d ApiListener#ListenerCoroutineProc(): get remote endpoint ASAP for logging
On incoming connection timeout we log the remote endpoint which isn't
available if it was already disconnected - an exception is thrown.  Get it
as long as we're still connected not to lose it, nor to get an exception.
2024-02-09 12:27:25 +01:00
Yonas Habteab 32531fe909 EventsHandler: Drop superfluous `CpuBoundWork` usage 2024-02-09 12:00:50 +01:00
Eric Lippmann c7293de91d IoEngine: Always log coroutine exception diagnostics
While analyzing a possible memory leak, we encountered several coroutine
exception messages, which unfortunately do not provide any information
about what exactly went wrong, as exception diagnostics were previously
only logged at the notice level.
2024-02-08 12:09:06 +01:00
Yonas Habteab 72266434df Drop redundant `CpuBoundWork` usages in `lib/remote` 2024-02-08 11:30:23 +01:00
Yonas Habteab e2793f1d88 Drop redundant `CpuBoundWork` usage in `JsonRpcConnection::Disconnect()`
Although there is locking involved here, it shoudln't take too long for
the thread to actually acquire it, since there aren't that many threads
dealing with endpoint clients concurrently. It's just wasting pointless
time trying to obtain a CPU slot.
2024-02-08 11:24:55 +01:00
Alexander Aleksandrovič Klimov e9fcbf400f
Merge pull request #9966 from Icinga/Al2Klimov-patch-3
HttpServerConnection: remove duplicate ")" from a log message
2024-01-18 10:46:51 +01:00
Alexander A. Klimov d48b369554 Reset all signal handlers of child processes
... not to disturb check plugins.

refs #6912
2024-01-17 12:25:59 +01:00
Alexander Aleksandrovič Klimov 966b46e808
Merge pull request #9965 from Icinga/http-request-time
HttpServerConnection: log request processing time as well
2024-01-17 11:30:33 +01:00
Julian Brost b1fe15f694
Merge pull request #9962 from Icinga/influx-disk-9948
Influx DB: truncate timestamps to whole seconds to save disk space
2024-01-17 08:50:16 +01:00
Alexander A. Klimov b6874cc8d4 HttpServerConnection: log request processing time as well 2024-01-16 17:52:07 +01:00
Alexander Aleksandrovič Klimov 6a4cb5c12c
HttpServerConnection: remove duplicate ")" from a log message
The commit 5c32a5a7dc, which introduced it, clearly shows that the other ")" already existed legitimately.
2024-01-16 16:31:00 +01:00
Alexander A. Klimov cc9db3756f Revert "Influx DB: don't unneccessarily truncate timestamps to whole seconds"
This reverts commit eaa3cd83ad.
2024-01-16 12:19:48 +01:00
Alexander A. Klimov fc5b1178c6 Revert "Remove no-op InfluxDB URL param"
This reverts commit 21f548d3c0.
2024-01-16 12:19:47 +01:00
Alexander Aleksandrovič Klimov 28b2db8446
Merge pull request #9851 from Icinga/Al2Klimov-patch-3
Make ObjectImpl<Logger>#GetSeverity() non-virtual
2023-12-22 12:44:51 +01:00
Alexander Aleksandrovič Klimov 6c03598678
Merge pull request #9896 from Icinga/provide-cancel_time-where-has_been_cancelled-may-be-1
Disallow triggering a cancelled downtime, but provide cancel_time in Icinga DB downtime history where has_been_cancelled may be 1
2023-12-20 10:03:09 +01:00
Alexander Aleksandrovič Klimov 949d983a76
Merge pull request #9895 from Icinga/targeted-api-filter
FilterUtility::GetFilterTargets(): don't run filter for specific object(s) for all objects
2023-12-19 15:18:41 +01:00
Alexander Aleksandrovič Klimov 8b2e28a869
Merge pull request #9891 from Icinga/renew-the-ca-9890
ApiListener#Start(): auto-renew CA on its owner
2023-12-19 14:57:47 +01:00
Alexander Aleksandrovič Klimov 96cfc4abe8
Merge pull request #9887 from Icinga/argument-list-too-long-9340
PluginNotificationTask::ScriptFunc(): on Linux truncate output and comment
2023-12-19 14:36:57 +01:00
Alexander A. Klimov 175153ce6a PluginNotificationTask::ScriptFunc(): on Linux truncate output and comment
not to run into an exec(3) error E2BIG due to a too long argument.
This sends a notification with truncated output instead of not sending.
2023-12-19 12:21:03 +01:00
Alexander A. Klimov 966216f4ba RequestCertificateHandler(): also renew if CA needs a renewal
and a newer one is available.
2023-12-18 15:28:11 +01:00
Alexander A. Klimov 551c3afa60 CertificateToString(): allow raw pointer input 2023-12-18 15:28:11 +01:00
Alexander A. Klimov bc778116e9 ApiListener#Start(): auto-renew CA on its owner
otherwise it would expire.
2023-12-18 15:28:11 +01:00
Alexander A. Klimov 36a08b0497 ApiListener#RenewCert(): enable optional CA creation 2023-12-18 15:28:11 +01:00
Alexander A. Klimov 7b55df6f11 CreateCertIcingaCA(EVP_PKEY*, X509_NAME*): enable optional CA creation 2023-12-18 15:28:11 +01:00
Alexander Aleksandrovič Klimov 953eeba061
Merge pull request #9893 from Icinga/do-not-re-notify-if-filtered-states-don-t-change-4503
Discard likely duplicate problem notifications via Notification#last_notified_state_per_user
2023-12-13 16:13:28 +01:00
Alexander A. Klimov ecfc9033b0 FilterUtility::GetFilterTargets(): don't run filter for specific object(s) for all objects 2023-12-13 16:02:50 +01:00
Alexander A. Klimov 15191bcd74 ApplyRule::GetTarget*s(): support constant strings from variables
in addition to literal strings. This is for sandboxed filters with some
variables pre-set by the caller. They're "constant" in that scope, too.
2023-12-13 16:02:50 +01:00
Alexander A. Klimov a04cef1890 Introduce DictExpression#GetExpressions() 2023-12-13 16:02:50 +01:00
Alexander A. Klimov 8bcae97ecc Introduce Dictionary#GetRef() 2023-12-13 16:02:50 +01:00
Alexander A. Klimov 97cd05db7a Notification#BeginExecuteNotification(): on recovery clear last_notified_state_per_user 2023-12-13 13:21:22 +01:00
Alexander A. Klimov 44e9c6f40d Notification#BeginExecuteNotification(): discard likely duplicate problem notifications 2023-12-13 13:21:19 +01:00
Alexander A. Klimov 74f52c6fcd Introduce IsCaUptodate() by splitting IsCertUptodate() 2023-12-13 12:08:34 +01:00
Julian Brost 871fa67b52
Merge pull request #9885 from Icinga/renegotiation 2023-12-12 17:38:09 +01:00
Alexander A. Klimov 2cff763295 Cluster-sync Notification#last_notified_state_per_user 2023-12-12 15:29:50 +01:00
Alexander A. Klimov b25ba7a316 Notification#BeginExecuteNotification(): track state change notifications 2023-12-07 12:43:30 +01:00
Julian Brost d2a7117007
Merge pull request #9899 from Icinga/icinga2-crashes-silently-9897
IcingaDB#SendConfigDelete(): fix missing nullptr check before deref
2023-11-21 11:03:28 +01:00
Alexander Aleksandrovič Klimov 7fc7d054af
Merge pull request #9841 from WuerthPhoenix/fix-9840-lock-console-api-during-reload 2023-11-21 10:36:26 +01:00
Alexander A. Klimov 7174dc864d IcingaDB#SendConfigDelete(): fix missing nullptr check before deref 2023-11-10 17:43:33 +01:00
Alexander A. Klimov 9aaa9901bd Icinga DB downtime history: provide cancel_time where has_been_cancelled may be 1
The table sla_history_downtime requires a downtime_end.
The Go daemon takes the cancel_time if has_been_cancelled is 1.
So we must supply a cancel_time whereever has_been_cancelled is 1.
Otherwise the Go daemon can't process some entries.
2023-11-08 15:22:39 +01:00
Alexander A. Klimov 7ce9457a4a Disable TLS renegotiation
The API doesn't need it and a customer's security scanner
is afraid of a potential DoS attack vector.
2023-11-06 18:46:37 +01:00
Theo Buehler 1f06589f7a Remove dead code in GetSignatureAlgorithm()
This code was added in commit 548eb93 and never did anything useful.
Using X509_get_signature_nid() or its expanded version in the pre-1.1
branch is the correct way of retrieving the signature algorithm of a
certificate.

CLA: trivial
2023-10-20 18:55:44 +02:00
Julian Brost bba6a76f4a
Merge pull request #9853 from Icinga/GelfWriter-m_StreamMutex
GelfWriter: protect m_Stream via m_WorkQueue, not ObjectLock(this)
2023-09-07 11:46:38 +02:00
Alexander Aleksandrovič Klimov e5d988a2fe
Merge pull request #7799 from Icinga/bugfix/file-end
Fix file endings
2023-08-25 11:06:19 +02:00
Alexander A. Klimov 4ee10a6c20 GelfWriter: protect m_Stream via m_WorkQueue, not ObjectLock(this)
On shutdown or HA re-connect ConfigObject#SetAuthority(false) is called which
does ObjectLock(this) and ConfigObject#Pause(). GelfWriter#Pause(), with the
above ObjectLock, calls m_WorkQueue.Join(). But items inside that also doing
ObjectLock(this) cause a deadlock.
2023-08-24 17:48:09 +02:00
Alexander Aleksandrovič Klimov 993c9b742d
Make ObjectImpl<Logger>#GetSeverity() non-virtual
After all it's not overridden.
2023-08-15 13:03:31 +02:00
Mattia Codato 41e21cb8cf Prevent calls to command API while the configuration is reloading.
Fixes #9840
2023-08-09 08:45:04 +02:00
Alexander A. Klimov 1308ad62af Stream#Read(): remove de facto unused param allow_partial
The only caller passes true, so no one forbids partial reads (even implicitly).
All usages in the implementation just assert it being true (allowed).
2023-07-13 16:55:48 +02:00
Alexander Aleksandrovič Klimov 1af5109ad3
Merge pull request #9734 from Icinga/remove-unused-stream-peek-
Remove unused Stream#Peek()
2023-07-13 16:52:29 +02:00
Alexander A. Klimov 8f8a6ee2a0 Application::m_LastReloadFailed: if double isn't always lock free, use uint32_t
which will overflow in 2106, not 2038.
This fixes a compile failure on 32-bit Raspbian.
2023-07-10 10:51:02 +02:00
Alexander Aleksandrovič Klimov 000a776dfb
Built-in check command: ifw-api (#9062) 2023-07-06 14:18:21 +02:00
Julian Brost 26a75f8a6f
Merge pull request #9812 from Icinga/support-elasticsearch-8-0-9251
ElasticsearchWriter: switch to v7+ URL schema to support v8
2023-07-05 10:15:10 +02:00
Julian Brost fe13b96226
Merge pull request #9809 from Icinga/reevaluate-and-update-default-tls-cipher-list-9808
Copy and paste global default TLS cipher set from ssl-config.mozilla.org
2023-07-03 19:13:10 +02:00
Alexander A. Klimov 617dda61fb Re-order global default TLS cipher list to prefer AES256 over AES128 2023-07-03 15:36:11 +02:00
Alexander A. Klimov 4c2e59a690 ElasticsearchWriter: switch to v7+ URL schema to support v8
and OpenSearch 2. This breaks the EOL v5 and v6.
2023-07-03 14:43:45 +02:00
Julian Brost 70d6b6e424
Merge pull request #9810 from Icinga/Al2Klimov-patch-8
ElasticsearchWriter#Pause(): call Flush() only once
2023-06-30 17:21:16 +02:00
Alexander Aleksandrovič Klimov 076eb59443
ElasticsearchWriter#Pause(): lock m_DataBufferMutex during Flush()
just to be sure regarding race conditions.
2023-06-30 14:57:18 +02:00
Julian Brost a2e05f89e8 Enable built-in OpenSSL DH parameters to allow DHE TLS ciphers
Non-ECC DHE ciphers in the `cipher_list` attribute of `ApiListener` (the
default value includes these) had no effect as no DH parameters were available
and therefore the server wouldn't offer these ciphers. OpenSSL provides
built-in DH parameters starting from version 1.1.0, however, these have to be
enables explicitly using the `SSL_CTX_set_dh_auto()` function. This commit does
so and thereby makes it possible to establish a connection to an Icinga 2
server using a DHE cipher.
2023-06-29 12:06:26 +02:00
Alexander Aleksandrovič Klimov d5e6ecec8a
ElasticsearchWriter#Pause(): call Flush() only once
The first Flush() is redundant and may access m_DataBuffer at the same time as some Flush() in m_WorkQueue (race condition) which isn't joined, yet.
2023-06-29 10:42:12 +02:00
Alexander A. Klimov 2e053b0e06 Copy and paste global default TLS cipher set from ssl-config.mozilla.org
which got more secure by now, but still overlaps with v2.13.x' set.
2023-06-28 14:49:08 +02:00
Julian Brost a2926b8604
Merge pull request #9794 from Icinga/round-notification-times-begin-end-not-to-crash-go-daemon
IcingaDB::PrepareObject(): round Notification#times.{begin,end} not to crash Go daemon
2023-06-27 17:08:41 +02:00
Alexander A. Klimov dccb678882 IcingaDB::PrepareObject(): cut off (null) negative Notification#times.{begin,end} not to crash Go daemon
At least our PostgreSQL schema enforces positive values.
2023-06-27 12:58:08 +02:00
Alexander A. Klimov 415b810abf IcingaDB::PrepareObject(): round Notification#times.{begin,end} not to crash Go daemon
The latter expects ints, not floats - not to mention strings.
Luckily Icinga already enforces numeric strings so that we can cast it to number.
2023-06-27 12:53:08 +02:00
Julian Brost 9cf519316e
Merge pull request #9805 from Icinga/checkcommand-timeout-0-crashes-icinga-db-daemon-9804
IcingaDB::PrepareObject(): cut off (0) negative Command#timeout for Redis
2023-06-27 10:45:02 +02:00
Julian Brost c08d3beeb1
Merge pull request #9785 from Icinga/Al2Klimov-patch-8
Icinga DB: also write ConfigObject#original_attributes into Redis
2023-06-27 10:24:41 +02:00
Julian Brost bd11bc2eb4
Merge pull request #9793 from Icinga/unmarshal-number-42-5-into-go-struct-field-notification-notification_interval
IcingaDB::PrepareObject(): round Notification#interval and limit it to >=0
2023-06-27 10:12:13 +02:00
Alexander A. Klimov d641a3c799 IcingaDB::PrepareObject(): cut off (0) negative Command#timeout for Redis
not to crash the Go daemon which expects positive values there.
2023-06-26 15:36:47 +02:00
Julian Brost 5350aa3c72
Merge pull request #9792 from Icinga/icingadb-conversion-of-strings-to-number-types-to-avoid-crashes-9791
IcingaDB::PrepareObject(): convert non-null Checkable#check_timeout to number
2023-06-26 15:03:21 +02:00
Alexander A. Klimov 273aa6f997 IcingaDB::PrepareObject(): round Notification#interval and limit it to >=0
otherwise, e.g. with -42.5, the Go daemon crashes. It expects uints there.
2023-06-19 12:46:40 +02:00
Alexander A. Klimov 9f08bad395 IcingaDB::PrepareObject(): convert non-null Checkable#check_timeout to number
and, in case of null, fall back to Checkable#check_command.timeout, just like
IcingaDB#SerializeState(). Otherwise the Go daemon crashes. It expects a number.
2023-06-15 12:29:42 +02:00
Alexander A. Klimov 1587431945 POST /v1/objects: allow array of attrs to undo modifications of 2023-06-13 16:40:33 +02:00
Alexander A. Klimov 385fe2fd76 Icinga DB: also write ConfigObject#original_attributes into Redis
for the case the Go daemon decides to sync them into DB.
2023-06-12 12:53:25 +02:00
Julian Brost 7c381ae12f
Merge pull request #9779 from Icinga/macroprocessor-resolvemacro-quasi-cv-object-icingaapplication
MacroProcessor::ResolveMacro(): treat quasi-CV-object IcingaApplication as real CV-object
2023-05-31 20:41:31 +02:00
Alexander A. Klimov a9c80ffb2e MacroProcessor::ResolveMacro(): treat quasi-CV-object IcingaApplication as real CV-object
As MacroProcessor checked just for CustomVarObject base class, but
IcingaApplication provided the vars attribute by itself, it had to also
resolve CV macros by itself. That logic diverged from MacroProcessor so that
macros inside IcingaApplication CVs weren't resolved. Until now.
2023-05-31 16:35:09 +02:00
Julian Brost 8a42c3bf18
Merge pull request #9775 from Icinga/icingadb-service-crashes-on-negative-downtime-duration-or-end-before-start-9774
Icinga DB: don't write negative Downtime durations into Redis
2023-05-31 11:37:42 +02:00
Alexander A. Klimov 75eaa81c06 Icinga DB: don't write negative Downtime durations into Redis
via `std::max(0, x)` not to crash the Go daemon which can't handle such.
2023-05-30 17:56:03 +02:00
Julian Brost b0899d9ab4
Merge pull request #8429 from Icinga/bugfix/last-reload-attempt-failed-8428
Share "Last reload attempt failed" time across Icinga process tree on *nix
2023-05-30 11:42:21 +02:00
Julian Brost d871c5c837
Merge pull request #9772 from Icinga/icinga-db-feature-should-normalize-command-arguments-required-skip_key-repeat_key-to-boolean-9576
Icinga DB feature: normalize *Command.arguments[*].{required,skip_key…
2023-05-25 11:54:01 +02:00
Alexander A. Klimov ad618e9716 Icinga DB feature: normalize *Command.arguments[*].{required,skip_key,repeat_key} to boolean
At the moment, the Icinga DB feature will use that value as-is and
serialize it to JSON, resulting in a crash in Icinga DB down the road
because it expects a boolean.
2023-05-24 16:04:14 +02:00
Julian Brost 2470e930eb
Merge pull request #9643 from Icinga/hardware_concurrency
Always use Configuration#Concurrency, not `std:🧵:hardware_concurrency()`
2023-05-23 19:23:14 +02:00
Alexander A. Klimov 3fae41ef22 Restart thread pool after freezing Configuration
The user (-D) or we could have changed Configuration.Concurrency,
so correct the thread pool's thread amount.
2023-05-23 14:41:35 +02:00
Julian Brost 0e25644151
Merge pull request #8969 from Icinga/bugfix/perfdata-dont-get-parsed-correctly-8912
PluginUtility: Fix PerfData parsing for values separated with multiple spaces
2023-05-22 17:16:31 +02:00
Alexander A. Klimov 9376a311ea Fix file endings
git ls-files -z \
	|grep -zEe '^lib/' \
	|grep -zEe '\.[ch]pp$' \
	|xargs -0 perl -p0i -e 's/\n*(?!(?:.|\n))/\n/'
2023-05-17 18:05:13 +02:00
Alexander A. Klimov 32eb1680f7 Configuration.Concurrency: default to 1 until Configuration freeze
not to start many threads before the user could override their amount (-D).
2023-05-11 16:59:47 +02:00
Alexander A. Klimov 8fb5d53118 Track Configuration.Concurrency modifications 2023-05-11 15:41:35 +02:00
Alexander A. Klimov 5c330e9d4f Share "Last reload attempt failed" time across Icinga process tree on *nix
... as only the umbrella process knows that time,
but the icinga check running in the main process also needs to know it.

refs #8428
2023-05-08 14:42:21 +02:00
Julian Brost eca8890d49
Merge pull request #9718 from Icinga/acknowledgement-sync-between-masters-are-not-working-9652
Checkable#ProcessCheckResult(): only clean up ack comments older than check result
2023-05-05 15:29:38 +02:00
Julian Brost af9d67b262
Merge pull request #9726 from Icinga/43624b
Remove -and notify- expired downtimes immediately, not every 60s II
2023-05-02 11:25:03 +02:00
Alexander A. Klimov 58b788cd51 Downtime#Start(): trigger flexible downtimes not earlier than fixed ones
the last state change could be a long time ago. If it's longer than
the new downtime's duration, the downtime expires immediately.

trigger time + duration < now
2023-04-18 16:55:32 +02:00
Julian Brost 8238ec0d96
Merge pull request #9725 from Icinga/operation_aborted-shutDownIfNeeded.Cancel
ApiListener#NewClientHandlerInternal(): on basic_socket#cancel() (due to timeout) don't ssl::stream#async_shutdown()
2023-04-17 12:21:21 +02:00
Alexander A. Klimov 0ac1cd1ecb Rename Downtime::DowntimesExpireTimerHandler()
to actually reflect its purpose.
2023-04-14 14:52:05 +02:00
Alexander A. Klimov 6adf2d19e4 Remove -and notify- expired downtimes immediately, not every 60s
Don't look for expired downtimes in a timer fired every 60s,
but fire one timer per downtime once at expire time.
2023-04-14 14:52:05 +02:00
Alexander A. Klimov ba7102cae3 Explicitly stop started timers and wait for them
before permitting their parent objects' destruction.
For the cases where the handlers have raw pointers to these objects.
2023-04-14 14:52:04 +02:00
Julian Brost 8228fae740
Merge pull request #8627 from WuerthPhoenix/bug/agent-cannot-update-executions-8616
Fix update execution message discarded. refs #8616
2023-04-13 19:29:49 +02:00
Julian Brost f505325ff9
Merge pull request #9445 from Icinga/9365
Disallow config modifications via API during reload
2023-04-13 17:11:58 +02:00
Mattia Codato c5c17928a6 Allow to exec command on endpoint where the checkable is not present but checkable has command_endpoint specified 2023-04-13 14:44:07 +02:00
Alexander A. Klimov 2ee776b5ab Disallow config modifications via API during reload
Once the new main process has read the config,
it misses subsequent modifications from the old process otherwise.
2023-04-12 14:45:40 +02:00
Alexander A. Klimov 64e000df56 Introduce ConfigObjects*Lock 2023-04-12 13:36:48 +02:00
Julian Brost 50018c1d2b
Merge pull request #8218 from efuss/redundancy_group
Introduce redundancy groups for Dependency Objects
2023-04-05 18:49:58 +02:00
Yonas Habteab 24d95e1178 PluginUtility: Fix PerfData don't get parsed correctly
The problem was that some PerfData labels contained several whitespace characters,
not just one, and therefore it was parsed incorrectly in `SplitPerfdata()`. I.e. the condition
in line 144 checks whether the first and last character is a normal quote, but since the
label can contain spaces at the beginning and at the end respectively, this caused the problems.

This PR fixes the problem by removing all occurring whitespace from the beginning and end,
before starting to parse the actual label.
2023-04-05 15:37:54 +02:00
Alexander A. Klimov a66ace7245 Introduce SharedMemory 2023-04-04 13:40:27 +02:00
Alexander A. Klimov c41e5fd05d Support multiple redundant Timer#Start() calls
so that only the first one changes l_AliveTimers (as in Timer#Stop()).
2023-04-04 10:35:22 +02:00
Alexander A. Klimov 298f3b1973 Timer: actually support non-periodic timers 2023-04-04 10:35:22 +02:00
Alexander A. Klimov 3933502739 Timer#Start(): don't unnecessarily unlock/lock l_TimerMutex
via new Timer#InternalRescheduleUnlocked()
2023-04-04 10:35:22 +02:00
Alexander A. Klimov 13b9cfda41 Timer::TimerThreadProc(): don't unnecessarily unlock and lock l_TimerMutex 2023-04-04 10:35:22 +02:00
Alexander A. Klimov 1badbab002 Timer::TimerThreadProc(): keep a Timer alive while it's running
to prevent the case: Timer callback destroys parent object -> destroys
Timer -> ~Timer() -> Stop(true) -> waits for the Timer callback to finish
-> deadlock.
2023-04-04 10:35:22 +02:00
Alexander A. Klimov 9b00c1c4dd Timer: drop unnecessary base class 2023-04-04 10:35:22 +02:00
Alexander A. Klimov 24681b30f6 Make Timer::Ptr a std::shared_ptr 2023-04-04 10:35:22 +02:00
Alexander A. Klimov 9ee4d08722 Make Timer#Timer() private to enforce Timer::Create() usage 2023-04-04 10:35:22 +02:00
Alexander A. Klimov 21b68455ce Use Timer::Create() instead of new Timer()
git ls-files -z |xargs -0 perl -pi -e 's/\bnew Timer\b/Timer::Create/g'

ex. in Timer::Create() itself.
2023-04-04 10:35:20 +02:00
Alexander A. Klimov bb1f574b69 Introduce factory method Timer::Create() 2023-04-04 10:35:10 +02:00
Alexander A. Klimov 35248b1b63 Code style 2023-04-03 13:39:08 +02:00
Alexander A. Klimov cc872dac1f Remove CheckResultReader which has been deprecated for 5 major versions 2023-04-03 11:39:21 +02:00
Julian Brost 7a7902cea7
Merge pull request #9715 from Icinga/StatusDataWriter
Remove StatusDataWriter which has been deprecated for 5 major versions
2023-03-31 12:32:43 +02:00
Julian Brost e87e1ea73f Freeze globals namespace during config load
This allows for a faster config load due to less locking required.

The change is slightly backwards-incompatible. Before, you could manipulate the
globals namespace at a later stage, but disallowing this feels reasonable for
the performance benefit alone (which especially shows on many-core machines).
Apart from that, it's doubtful if doing so is even useful at all as the DSL
provides no mechanism for you to synchronize your operations that may run in
parallel. The data structures itself are protected from race conditions, but
anything implemented on top of this may still be subject to race conditions.
And even if some user has a good reason for doing this, there's a feasible
workaround by creating your own namespace like globals.mutable and using that
instead.
2023-03-30 18:07:51 +02:00
Alexander A. Klimov 335688909b Document why Timer::TimerThreadProc() can use Timer members during Timer#~Timer() call 2023-03-29 18:04:19 +02:00
Alexander A. Klimov 78b4dc6509 Remove unused Stream#Peek() 2023-03-24 18:18:13 +01:00
Alexander A. Klimov 4c154f93dc ApiListener#NewClientHandlerInternal(): on basic_socket#cancel() (due to timeout) don't ssl::stream#async_shutdown()
If a connection hangs for too long in ApiListener#NewClientHandler(),
ApiListener#AddConnection()'s Timeout calls boost::asio::basic_socket#cancel()
on that connection to trigger an exception which unwinds
ApiListener#NewClientHandler(). Previously that unwind could trigger a Defer
which called boost::asio::ssl::stream#async_shutdown() which extended the hang.
2023-03-21 10:57:40 +01:00
Julian Brost 66b039df9c
Merge pull request #9497 from Icinga/9249
Application::Exit(): don't exit(), but _exit(), even in debug build mode
2023-03-10 16:04:54 +01:00
Alexander A. Klimov 6414fd19f5 Checkable#ProcessCheckResult(): only clean up ack comments older than check result
Normally if for some reason an ack comment still exists on a checkable not
acked anymore, still clean it up. But while replaying log config objects
incl. ack comments come before check results and acks. I.e. 1) ack comment,
2) DOWN check result and 3) ack. Not 1) DOWN check result, 2) ack and 3) ack
comment. So the checkable is temporarily not acked, but already has the ack
comment. In this case the DOWN check result which is older than the ack
comment shall not clean up the latter.
2023-03-03 15:48:34 +01:00
Alexander A. Klimov 4662d4477b Checkable#RemoveAckComments(): add optional comment entry time filter 2023-03-03 15:48:11 +01:00
Alexander A. Klimov dceb29c742 Checkable#RemoveCommentsByType(): remove redundant parameter 2023-03-03 11:53:02 +01:00
Mattia Codato 912fdb9700 Fix update execution message discarded
refs Icinga#8616
2023-03-02 17:50:39 +01:00
Alexander Aleksandrovič Klimov 55930c8042
ProcessSpawnImpl(): remove redundant _exit(128);
Now this if doesn’t _exit(128) by itself, but "return" to the outer if which immediately _exit(128)s.
2023-03-02 12:45:15 +01:00
Alexander A. Klimov bbf2e80002 Remove StatusDataWriter which has been deprecated for 5 major versions 2023-03-01 17:16:28 +01:00