[SIEM] Netflow Data always displays 0 bytes in / out #14212

StefanSa · 2019-10-12T12:05:28Z

Kibana version:
7.4
Elasticsearch version:
7.4
Filebeat version:
7.4
SIEM App always shows 0 bytes for Netflow data. Filebeat, however, transmits correct data.
This problem exists in general and not only in this example.

Screenshots (if relevant):

SIEM App:

Filebeat Netflow Dashboard:

elasticmachine · 2019-10-15T17:06:53Z

Pinging @elastic/siem (Team:SIEM)

kicker83 · 2019-10-22T09:36:26Z

Hi! same problem here, I'm getting unknown direction, 0 bytes and 0 packets. We are collecting Netflow from Cisco HSL (ASR and ISR SDWAN)

Thanks!

adriansr · 2019-10-22T10:10:10Z

Do your Netflow events contain source.bytes and source.packets? If that's the case, my guess is this will be fixed in 7.5 by #14111 .

kicker83 · 2019-10-22T10:15:05Z

In Wireshark I can see "FlowSet Length", but I'm not able to see a field named "source.bytes" as is.

Thanks!

adriansr · 2019-10-22T10:37:26Z

I mean the "source.bytes" field on the Elasticsearch document

kicker83 · 2019-10-22T10:40:20Z

That field is defined in Filebeat index pattern, but there is no value indexed in elasticsearch.

adriansr · 2019-10-22T13:04:10Z

Then you are having the problem fixed by #14111, it will be fixed in 7.5.0. You can try a snapshot build: https://beats-ci.elastic.co/job/elastic+beats+7.5+multijob-package-linux/lastSuccessfulBuild/gcsObjects/

kicker83 · 2019-10-22T13:23:36Z

Ok Adrian! I will give a try to the snapshot. I will confirm here if it works.

Thanks so much for your help :)

kicker83 · 2019-10-22T15:22:18Z

I tested the Filebeat snapshot, but I still don't have netflow bytes and network packets neither.

I cleaned index patterns in kibana and indexes in elasticsearch, and then I made a new setup in filebeat, but no luck.

Maybe some strange format in the netflow, I don't know :(

Anyway, thanks for your help!

adriansr · 2019-10-22T17:26:42Z

You're sure you're looking at events generated by 7.5.0?

Can you share one of those documents without "source.bytes" / "network.bytes", in JSON format? That will include the "raw" Netflow fields so will be useful to see if there's a mapping issue.

kicker83 · 2019-10-23T08:04:34Z

Hi Adrian, It seems to be in 7.5.0. In index management I can see this format "filebeat-7.5.0-2019.10.22-000001" and in the index pattern the field count has raised to 1786. Here is the JSON output from kibana: kibana_export.json And here is the JSON output from tshark. Its a network capture: tshark_netflow.json I hope this could help. Thanks for your help Regards El mar., 22 oct. 2019 a las 19:27, Adrian Serrano (<notifications@github.com>) escribió:

You're sure you're looking at events generated by 7.5.0? Can you share one of those documents without "source.bytes", in JSON format? That will include the "raw" Netflow fields so will be useful to see if there's a mapping issue. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://github.com/elastic/kibana/issues/48052?email_source=notifications&email_token=AETJEFU7UBGFBEQRCH7QBVLQP4ZXZA5CNFSM4JACTYY2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEB6RMEQ#issuecomment-545068562>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AETJEFXLINE6UKEW3DXWF4DQP4ZXZANCNFSM4JACTYYQ> .

[kibana_export.json.txt](https://github.com/elastic/kibana/files/3760969/kibana_export.json.txt) [tshark_netflow.json.txt](https://github.com/elastic/kibana/files/3760970/tshark_netflow.json.txt)

adriansr · 2019-10-23T15:52:46Z

Ok, the ES document looks good, it has all the fields that show up in the tshark dump.

The issue is that your netflow exporter (the device that is sending netflow to filebeat) is not adding fields for bytes/packet counters. It's only adding initiatorOctets and responderOctets, which are L4 payload bytes that we don't use.

To have source.bytes and packets you need to change its configuration so that the octetDeltaCount and packetDeltaCount are exported.

StefanSa · 2019-12-03T11:19:07Z

@adriansr
Problem still exists in v7.5.
Is there a chance that the problem will be solved this year?

StefanSa · 2019-12-06T16:08:59Z

@adriansr
any ETA here when the bug is fixed?

andrewkroh · 2019-12-10T00:11:54Z

Unidirectional flows are populating source.bytes and source.packets in Filebeat v7.5.0. The SIEM Network page will have one non-zero column for these unidirectional flow.

adriansr · 2019-12-10T07:52:57Z

@StefanSa as I said in a comment above, the problem here is that your Netflow templates are not exporting the necessary fields (octetDeltaCount and packetDeltaCount), you need to change the configuration on your netflow exporter so those fields are sent to Filebeat.

This is not a bug on our side.

StefanSa · 2019-12-10T09:54:29Z

@adriansr
Thank you for your answer.
Question, we are talking about IPFIX (octetDeltaCount and packetDeltaCount) or Netflow ?

adriansr · 2020-01-07T21:13:29Z

Turns out this is caused by Cisco reusing using fields IDs for different meanings in some devices.

Cisco's documentation for NetFlow states:

For the information on the field types with the numbers between 128 and 32768, please refer to the IANA registry of IPFIX information elements at http://www.iana.org/assignments/ipfix

Which means that fields ID 231 and 232 are the layer 4 bytes counters initiatorOctets and responderOctets.

But in the Cisco ASA Netflow docs those same field IDs are aparently used with a different meaning:

Field name	ID	Size	Description
NF_F_FWD_FLOW_DELTA_BYTES	231	4	The delta number of bytes from source to destination.
NF_F_REV_FLOW_DELTA_BYTES	232	4	The delta number of bytes from destination to source.

I think the correct fix is to stick with IPFIX definitions for fields and update our ECS conversion to use initiatorOctets and responderOctets when octetDeltaCount and packetDeltaCount are missing.

Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. Fixes elastic#14212

- Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes #14212

…#15449) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes elastic#14212 (cherry picked from commit c3a3604)

…#15554) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes #14212 (cherry picked from commit c3a3604)

…#15555) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes #14212 (cherry picked from commit c3a3604)

* [Filebeat] Fixes for NetFlow v9 devices from various vendors (#15449) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes #14212 * update settings for `decode_csv_fields` (#15249) (#15550) Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> * docs: updates to output config (#15272) * [Filebeat] Handle error message in handleS3Objects function (#15545) * Handle error message in handleS3Objects function * remove s3Context.Fail and use setError and done instead * Add changelog * Fix use of wrong fields in Cisco ASA dashboard (#15553) This dashboard wasn't updated after a couple of fields were renamed. Fixes: #15420 * Add test for publisher spool encode and decode. (#15534) * Add test for publisher queue encode and decode. * Run mage fmt. * Fixes from code review. * Fix convert processor conversion of string with leading zeros to integer (#15557) The conversion failed when for strings with leading zeroes and a decimal digit 8 or 9, as the underlying runtime function would try to parse that as an octal number. This is fixed by only allowing decimal and hex, which in turns makes the processor more aligned to its Elasticsearch counterpart. Fixes #15513 * New mage target: generate pkg file to test the manager (#15580) This PR adds a new mage target to Functionbeat named `buildPkgForFunction`. It generates the folder `pkg` with the functions to make testing the manager more comfortable during development. * Packetbeat TLS: Replace array fields with keyword (#15597) Use of `type: array` in some fields (which was inconsistent) causes those fields to be excluded from the template. This prevents pointing aliases to those fields, which we need in 7.6+. Setting those fields to `keyword` explicitly so that they are included in the template. Fixes #15588 * Add a pull request template providing valuable information when reviewing a PR (#15388) * Add a PR template that provides valuable information when reviewing a PR * Add CLA check * Fix typo * Address comments during review * SF: Fix typo * Add deprecation as PR type * Make it clear how to strike through in markdown * Add default configuration files to the checklist * [Metricbeat] Implement IBM MQ module (#15301) * Modify cockroachdb source * Define testdata * Do not publish ports * Update docs * mage fmt update * Describe containerized environment * Update CHANGELOG.next.asciidoc Co-Authored-By: Chris Mark <chrismarkou92@gmail.com> * Update data.json * Rename image * Update source after review * Filter ibmmq_ metrics * mage check * Fix: mage check * Don't expose port * Rename status to qmgr * Add subscriptions overview dashboard for IBM MQ module * Add calls, messages overview dashboard for IBM MQ module * Add screenshots * Fix: mage check * Fix: CHANGELOG * Add explanation * Fix: mage check Co-authored-by: Chris Mark <chrismarkou92@gmail.com> * Cleanup changelogs for master (#15617) * Cleanup changelogs for master * Remove extra header in CHANGELOG.asciidoc * [Metricbeat] Add lambda metricset in aws module (#15260) * Add lambda metricset * Adds missing imports (#15624) * [docs] Clarify privileges required for the writer role (#15604) * Mask password discovered via module autodiscover hint (#15616) * Mask password is string representation of config * Rename method * Adding unit test * Use const for module config password setting name * Using common.DebugString * Simplifying * Removing now-invalid unit test * Removing now-unnecessary const * Refactoring: moving debug-related var and func to common file * Refactoring: rename from black list to mask list * Implement fmt.Formatter for common.MapStr * Reintroduce debug statement * Make MarshalLogObject always filter MapStr object for logging purposes * Refactoring: renaming to be bit more generic * Forgot to add license header to new file * Fixing verb syntax * Update KQL to get estimated cost without dimension ServiceName (#15640) * Adding monitoring.cloud.* settings to reference files (#15648) * Adding monitoring.cloud.* settings to reference files * Missed winlogbeat somehow * Missed x-pack/winlogbeat * remove lablels (#15644) * Fix panic: don't send events if client is nil (#15568) * Fix panic: don't send events if client is nil * Use mutex * Add CHANGELOG entry * Rename changelog entry * Fix: changelog * Temporarily use specific logstash release * [Metricbeat] Add Istio mesh metricset (#15535) * [Metricbeat] Fix changelog (#15681) * Fix changelog * ci: use APM pipeline library (#15636) it uses APM pipeline library configured in the instance * AWS Lambda: downgrade Kibana dashboard (#15682) * AWS Lambda: downgrade Kibana dashboard * Downgrade other AWS dashboards * Log command error if setup dashboards fails * Another downgrade * Use github.com/godror/godror instead of goracle.v2 (#15683) From the README of goracle: > Goracle is deprecated because of naming (trademark) issues. From now on we are using github.com/godror/godror instead. * Move pdh query to shared location in order for new modules/metricsets to reuse (#15503) * Move pdh query to shared location * Update changelog * Fix make update * mage fmt * fix changelog * Remove datasource option from SQL module and add tests (#15686) Remove datasource option from SQL module. This option was intended to set the DSN of a database connection, and we were ignoring the hosts setting. In other SQL modules we are using the values in hosts as DSNs, do here the same for consistency. Host is redacted when we cannot parse it as it can contain passwords. StandardizeEvent is exposed in mbtest.Fetcher interface so we can more easily check contents of events in tests. Add integration tests of the module with MySQL and PostgreSQL. Add real data.json with data from MySQL and PostgreSQL. * [metricbeat] add service metricset to reference documentation (#15643) * add service metricset to ref docs * update xpack docs * [metricbeat] Add divide by zero check to docker/diskio (#15649) * add NaN check to docker/diskio * return 0 * Change joda style pattern to java style (#15695) since 7.0 elasticsearch is using java.time style patterns. YYYY becomes yyyy * [DOCS] Add missing config options to shared file (#15136) * [DOCS] Add missing config options to shared file * Add fixes from review * Run mage fmt update to fix build error * [Heartbeat] Support for multiple status codes #13595 (#15587) Allow for multiple status codes in config. Fixes #13595 * Add missing changelog entry for #15587 (#15721) * Update github.com/godror/godror to v0.10.4 (#15737) ## What does this PR do? This PR updates the dependency `github.com/godror/godror` to v0.10.4. ## Why is it important? Packaging of Metricbeat fails due to the issue in the `godror` version we are currently using. See more about the problem: godror/godror#8 * Collect normalized CPU percentages by default (#15729) * Collect normalized CPU percentages by default * Adding CHANGELOG entry * Updating x-pack/metricbeat * Fix: mage check * Detect Eclipse Public License Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com>

* Add MQTT input to Filebeat (#15287) * Inital commit for MQTT input * Improved naming and error handling * Improved naming and connection procedure * Merge "master" branch into "feature-mqtt-input" (#15745) * [Filebeat] Fixes for NetFlow v9 devices from various vendors (#15449) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes #14212 * update settings for `decode_csv_fields` (#15249) (#15550) Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> * docs: updates to output config (#15272) * [Filebeat] Handle error message in handleS3Objects function (#15545) * Handle error message in handleS3Objects function * remove s3Context.Fail and use setError and done instead * Add changelog * Fix use of wrong fields in Cisco ASA dashboard (#15553) This dashboard wasn't updated after a couple of fields were renamed. Fixes: #15420 * Add test for publisher spool encode and decode. (#15534) * Add test for publisher queue encode and decode. * Run mage fmt. * Fixes from code review. * Fix convert processor conversion of string with leading zeros to integer (#15557) The conversion failed when for strings with leading zeroes and a decimal digit 8 or 9, as the underlying runtime function would try to parse that as an octal number. This is fixed by only allowing decimal and hex, which in turns makes the processor more aligned to its Elasticsearch counterpart. Fixes #15513 * New mage target: generate pkg file to test the manager (#15580) This PR adds a new mage target to Functionbeat named `buildPkgForFunction`. It generates the folder `pkg` with the functions to make testing the manager more comfortable during development. * Packetbeat TLS: Replace array fields with keyword (#15597) Use of `type: array` in some fields (which was inconsistent) causes those fields to be excluded from the template. This prevents pointing aliases to those fields, which we need in 7.6+. Setting those fields to `keyword` explicitly so that they are included in the template. Fixes #15588 * Add a pull request template providing valuable information when reviewing a PR (#15388) * Add a PR template that provides valuable information when reviewing a PR * Add CLA check * Fix typo * Address comments during review * SF: Fix typo * Add deprecation as PR type * Make it clear how to strike through in markdown * Add default configuration files to the checklist * [Metricbeat] Implement IBM MQ module (#15301) * Modify cockroachdb source * Define testdata * Do not publish ports * Update docs * mage fmt update * Describe containerized environment * Update CHANGELOG.next.asciidoc Co-Authored-By: Chris Mark <chrismarkou92@gmail.com> * Update data.json * Rename image * Update source after review * Filter ibmmq_ metrics * mage check * Fix: mage check * Don't expose port * Rename status to qmgr * Add subscriptions overview dashboard for IBM MQ module * Add calls, messages overview dashboard for IBM MQ module * Add screenshots * Fix: mage check * Fix: CHANGELOG * Add explanation * Fix: mage check Co-authored-by: Chris Mark <chrismarkou92@gmail.com> * Cleanup changelogs for master (#15617) * Cleanup changelogs for master * Remove extra header in CHANGELOG.asciidoc * [Metricbeat] Add lambda metricset in aws module (#15260) * Add lambda metricset * Adds missing imports (#15624) * [docs] Clarify privileges required for the writer role (#15604) * Mask password discovered via module autodiscover hint (#15616) * Mask password is string representation of config * Rename method * Adding unit test * Use const for module config password setting name * Using common.DebugString * Simplifying * Removing now-invalid unit test * Removing now-unnecessary const * Refactoring: moving debug-related var and func to common file * Refactoring: rename from black list to mask list * Implement fmt.Formatter for common.MapStr * Reintroduce debug statement * Make MarshalLogObject always filter MapStr object for logging purposes * Refactoring: renaming to be bit more generic * Forgot to add license header to new file * Fixing verb syntax * Update KQL to get estimated cost without dimension ServiceName (#15640) * Adding monitoring.cloud.* settings to reference files (#15648) * Adding monitoring.cloud.* settings to reference files * Missed winlogbeat somehow * Missed x-pack/winlogbeat * remove lablels (#15644) * Fix panic: don't send events if client is nil (#15568) * Fix panic: don't send events if client is nil * Use mutex * Add CHANGELOG entry * Rename changelog entry * Fix: changelog * Temporarily use specific logstash release * [Metricbeat] Add Istio mesh metricset (#15535) * [Metricbeat] Fix changelog (#15681) * Fix changelog * ci: use APM pipeline library (#15636) it uses APM pipeline library configured in the instance * AWS Lambda: downgrade Kibana dashboard (#15682) * AWS Lambda: downgrade Kibana dashboard * Downgrade other AWS dashboards * Log command error if setup dashboards fails * Another downgrade * Use github.com/godror/godror instead of goracle.v2 (#15683) From the README of goracle: > Goracle is deprecated because of naming (trademark) issues. From now on we are using github.com/godror/godror instead. * Move pdh query to shared location in order for new modules/metricsets to reuse (#15503) * Move pdh query to shared location * Update changelog * Fix make update * mage fmt * fix changelog * Remove datasource option from SQL module and add tests (#15686) Remove datasource option from SQL module. This option was intended to set the DSN of a database connection, and we were ignoring the hosts setting. In other SQL modules we are using the values in hosts as DSNs, do here the same for consistency. Host is redacted when we cannot parse it as it can contain passwords. StandardizeEvent is exposed in mbtest.Fetcher interface so we can more easily check contents of events in tests. Add integration tests of the module with MySQL and PostgreSQL. Add real data.json with data from MySQL and PostgreSQL. * [metricbeat] add service metricset to reference documentation (#15643) * add service metricset to ref docs * update xpack docs * [metricbeat] Add divide by zero check to docker/diskio (#15649) * add NaN check to docker/diskio * return 0 * Change joda style pattern to java style (#15695) since 7.0 elasticsearch is using java.time style patterns. YYYY becomes yyyy * [DOCS] Add missing config options to shared file (#15136) * [DOCS] Add missing config options to shared file * Add fixes from review * Run mage fmt update to fix build error * [Heartbeat] Support for multiple status codes #13595 (#15587) Allow for multiple status codes in config. Fixes #13595 * Add missing changelog entry for #15587 (#15721) * Update github.com/godror/godror to v0.10.4 (#15737) ## What does this PR do? This PR updates the dependency `github.com/godror/godror` to v0.10.4. ## Why is it important? Packaging of Metricbeat fails due to the issue in the `godror` version we are currently using. See more about the problem: godror/godror#8 * Collect normalized CPU percentages by default (#15729) * Collect normalized CPU percentages by default * Adding CHANGELOG entry * Updating x-pack/metricbeat * Fix: mage check * Detect Eclipse Public License Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com> * [Filebeat] Refactor mqtt input (#16014) * Refactor mqtt input * Fix: comment * Add unit tests * Test: input run * Fix Test: run and stop * Test: backoff * Adjust code after review * MQTT: update docs (#16152) * MQTT: add integration test (#16143) * Create mosquitto image * MQTT input: add integration test * Fix * Verify connectivity * Fix * Fix: mage check * Fix * Fix * Fix: remove global var * Update changelog * Fix: regenerate notice file * Remove unused dependency * Fix: zero qos * Wait asynchronously for client being disconnected Co-authored-by: Felix <felix.roessel@elastic.co> Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com>

* Add MQTT input to Filebeat (elastic#15287) * Inital commit for MQTT input * Improved naming and error handling * Improved naming and connection procedure * Merge "master" branch into "feature-mqtt-input" (elastic#15745) * [Filebeat] Fixes for NetFlow v9 devices from various vendors (elastic#15449) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes elastic#14212 * update settings for `decode_csv_fields` (elastic#15249) (elastic#15550) Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> * docs: updates to output config (elastic#15272) * [Filebeat] Handle error message in handleS3Objects function (elastic#15545) * Handle error message in handleS3Objects function * remove s3Context.Fail and use setError and done instead * Add changelog * Fix use of wrong fields in Cisco ASA dashboard (elastic#15553) This dashboard wasn't updated after a couple of fields were renamed. Fixes: elastic#15420 * Add test for publisher spool encode and decode. (elastic#15534) * Add test for publisher queue encode and decode. * Run mage fmt. * Fixes from code review. * Fix convert processor conversion of string with leading zeros to integer (elastic#15557) The conversion failed when for strings with leading zeroes and a decimal digit 8 or 9, as the underlying runtime function would try to parse that as an octal number. This is fixed by only allowing decimal and hex, which in turns makes the processor more aligned to its Elasticsearch counterpart. Fixes elastic#15513 * New mage target: generate pkg file to test the manager (elastic#15580) This PR adds a new mage target to Functionbeat named `buildPkgForFunction`. It generates the folder `pkg` with the functions to make testing the manager more comfortable during development. * Packetbeat TLS: Replace array fields with keyword (elastic#15597) Use of `type: array` in some fields (which was inconsistent) causes those fields to be excluded from the template. This prevents pointing aliases to those fields, which we need in 7.6+. Setting those fields to `keyword` explicitly so that they are included in the template. Fixes elastic#15588 * Add a pull request template providing valuable information when reviewing a PR (elastic#15388) * Add a PR template that provides valuable information when reviewing a PR * Add CLA check * Fix typo * Address comments during review * SF: Fix typo * Add deprecation as PR type * Make it clear how to strike through in markdown * Add default configuration files to the checklist * [Metricbeat] Implement IBM MQ module (elastic#15301) * Modify cockroachdb source * Define testdata * Do not publish ports * Update docs * mage fmt update * Describe containerized environment * Update CHANGELOG.next.asciidoc Co-Authored-By: Chris Mark <chrismarkou92@gmail.com> * Update data.json * Rename image * Update source after review * Filter ibmmq_ metrics * mage check * Fix: mage check * Don't expose port * Rename status to qmgr * Add subscriptions overview dashboard for IBM MQ module * Add calls, messages overview dashboard for IBM MQ module * Add screenshots * Fix: mage check * Fix: CHANGELOG * Add explanation * Fix: mage check Co-authored-by: Chris Mark <chrismarkou92@gmail.com> * Cleanup changelogs for master (elastic#15617) * Cleanup changelogs for master * Remove extra header in CHANGELOG.asciidoc * [Metricbeat] Add lambda metricset in aws module (elastic#15260) * Add lambda metricset * Adds missing imports (elastic#15624) * [docs] Clarify privileges required for the writer role (elastic#15604) * Mask password discovered via module autodiscover hint (elastic#15616) * Mask password is string representation of config * Rename method * Adding unit test * Use const for module config password setting name * Using common.DebugString * Simplifying * Removing now-invalid unit test * Removing now-unnecessary const * Refactoring: moving debug-related var and func to common file * Refactoring: rename from black list to mask list * Implement fmt.Formatter for common.MapStr * Reintroduce debug statement * Make MarshalLogObject always filter MapStr object for logging purposes * Refactoring: renaming to be bit more generic * Forgot to add license header to new file * Fixing verb syntax * Update KQL to get estimated cost without dimension ServiceName (elastic#15640) * Adding monitoring.cloud.* settings to reference files (elastic#15648) * Adding monitoring.cloud.* settings to reference files * Missed winlogbeat somehow * Missed x-pack/winlogbeat * remove lablels (elastic#15644) * Fix panic: don't send events if client is nil (elastic#15568) * Fix panic: don't send events if client is nil * Use mutex * Add CHANGELOG entry * Rename changelog entry * Fix: changelog * Temporarily use specific logstash release * [Metricbeat] Add Istio mesh metricset (elastic#15535) * [Metricbeat] Fix changelog (elastic#15681) * Fix changelog * ci: use APM pipeline library (elastic#15636) it uses APM pipeline library configured in the instance * AWS Lambda: downgrade Kibana dashboard (elastic#15682) * AWS Lambda: downgrade Kibana dashboard * Downgrade other AWS dashboards * Log command error if setup dashboards fails * Another downgrade * Use github.com/godror/godror instead of goracle.v2 (elastic#15683) From the README of goracle: > Goracle is deprecated because of naming (trademark) issues. From now on we are using github.com/godror/godror instead. * Move pdh query to shared location in order for new modules/metricsets to reuse (elastic#15503) * Move pdh query to shared location * Update changelog * Fix make update * mage fmt * fix changelog * Remove datasource option from SQL module and add tests (elastic#15686) Remove datasource option from SQL module. This option was intended to set the DSN of a database connection, and we were ignoring the hosts setting. In other SQL modules we are using the values in hosts as DSNs, do here the same for consistency. Host is redacted when we cannot parse it as it can contain passwords. StandardizeEvent is exposed in mbtest.Fetcher interface so we can more easily check contents of events in tests. Add integration tests of the module with MySQL and PostgreSQL. Add real data.json with data from MySQL and PostgreSQL. * [metricbeat] add service metricset to reference documentation (elastic#15643) * add service metricset to ref docs * update xpack docs * [metricbeat] Add divide by zero check to docker/diskio (elastic#15649) * add NaN check to docker/diskio * return 0 * Change joda style pattern to java style (elastic#15695) since 7.0 elasticsearch is using java.time style patterns. YYYY becomes yyyy * [DOCS] Add missing config options to shared file (elastic#15136) * [DOCS] Add missing config options to shared file * Add fixes from review * Run mage fmt update to fix build error * [Heartbeat] Support for multiple status codes elastic#13595 (elastic#15587) Allow for multiple status codes in config. Fixes elastic#13595 * Add missing changelog entry for elastic#15587 (elastic#15721) * Update github.com/godror/godror to v0.10.4 (elastic#15737) ## What does this PR do? This PR updates the dependency `github.com/godror/godror` to v0.10.4. ## Why is it important? Packaging of Metricbeat fails due to the issue in the `godror` version we are currently using. See more about the problem: godror/godror#8 * Collect normalized CPU percentages by default (elastic#15729) * Collect normalized CPU percentages by default * Adding CHANGELOG entry * Updating x-pack/metricbeat * Fix: mage check * Detect Eclipse Public License Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com> * [Filebeat] Refactor mqtt input (elastic#16014) * Refactor mqtt input * Fix: comment * Add unit tests * Test: input run * Fix Test: run and stop * Test: backoff * Adjust code after review * MQTT: update docs (elastic#16152) * MQTT: add integration test (elastic#16143) * Create mosquitto image * MQTT input: add integration test * Fix * Verify connectivity * Fix * Fix: mage check * Fix * Fix * Fix: remove global var * Update changelog * Fix: regenerate notice file * Remove unused dependency * Fix: zero qos * Wait asynchronously for client being disconnected Co-authored-by: Felix <felix.roessel@elastic.co> Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com> (cherry picked from commit 56c93e5)

…6284) * Filebeat: Merge "mqtt" input to master (#16204) * Add MQTT input to Filebeat (#15287) * Inital commit for MQTT input * Improved naming and error handling * Improved naming and connection procedure * Merge "master" branch into "feature-mqtt-input" (#15745) * [Filebeat] Fixes for NetFlow v9 devices from various vendors (#15449) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes #14212 * update settings for `decode_csv_fields` (#15249) (#15550) Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> * docs: updates to output config (#15272) * [Filebeat] Handle error message in handleS3Objects function (#15545) * Handle error message in handleS3Objects function * remove s3Context.Fail and use setError and done instead * Add changelog * Fix use of wrong fields in Cisco ASA dashboard (#15553) This dashboard wasn't updated after a couple of fields were renamed. Fixes: #15420 * Add test for publisher spool encode and decode. (#15534) * Add test for publisher queue encode and decode. * Run mage fmt. * Fixes from code review. * Fix convert processor conversion of string with leading zeros to integer (#15557) The conversion failed when for strings with leading zeroes and a decimal digit 8 or 9, as the underlying runtime function would try to parse that as an octal number. This is fixed by only allowing decimal and hex, which in turns makes the processor more aligned to its Elasticsearch counterpart. Fixes #15513 * New mage target: generate pkg file to test the manager (#15580) This PR adds a new mage target to Functionbeat named `buildPkgForFunction`. It generates the folder `pkg` with the functions to make testing the manager more comfortable during development. * Packetbeat TLS: Replace array fields with keyword (#15597) Use of `type: array` in some fields (which was inconsistent) causes those fields to be excluded from the template. This prevents pointing aliases to those fields, which we need in 7.6+. Setting those fields to `keyword` explicitly so that they are included in the template. Fixes #15588 * Add a pull request template providing valuable information when reviewing a PR (#15388) * Add a PR template that provides valuable information when reviewing a PR * Add CLA check * Fix typo * Address comments during review * SF: Fix typo * Add deprecation as PR type * Make it clear how to strike through in markdown * Add default configuration files to the checklist * [Metricbeat] Implement IBM MQ module (#15301) * Modify cockroachdb source * Define testdata * Do not publish ports * Update docs * mage fmt update * Describe containerized environment * Update CHANGELOG.next.asciidoc Co-Authored-By: Chris Mark <chrismarkou92@gmail.com> * Update data.json * Rename image * Update source after review * Filter ibmmq_ metrics * mage check * Fix: mage check * Don't expose port * Rename status to qmgr * Add subscriptions overview dashboard for IBM MQ module * Add calls, messages overview dashboard for IBM MQ module * Add screenshots * Fix: mage check * Fix: CHANGELOG * Add explanation * Fix: mage check Co-authored-by: Chris Mark <chrismarkou92@gmail.com> * Cleanup changelogs for master (#15617) * Cleanup changelogs for master * Remove extra header in CHANGELOG.asciidoc * [Metricbeat] Add lambda metricset in aws module (#15260) * Add lambda metricset * Adds missing imports (#15624) * [docs] Clarify privileges required for the writer role (#15604) * Mask password discovered via module autodiscover hint (#15616) * Mask password is string representation of config * Rename method * Adding unit test * Use const for module config password setting name * Using common.DebugString * Simplifying * Removing now-invalid unit test * Removing now-unnecessary const * Refactoring: moving debug-related var and func to common file * Refactoring: rename from black list to mask list * Implement fmt.Formatter for common.MapStr * Reintroduce debug statement * Make MarshalLogObject always filter MapStr object for logging purposes * Refactoring: renaming to be bit more generic * Forgot to add license header to new file * Fixing verb syntax * Update KQL to get estimated cost without dimension ServiceName (#15640) * Adding monitoring.cloud.* settings to reference files (#15648) * Adding monitoring.cloud.* settings to reference files * Missed winlogbeat somehow * Missed x-pack/winlogbeat * remove lablels (#15644) * Fix panic: don't send events if client is nil (#15568) * Fix panic: don't send events if client is nil * Use mutex * Add CHANGELOG entry * Rename changelog entry * Fix: changelog * Temporarily use specific logstash release * [Metricbeat] Add Istio mesh metricset (#15535) * [Metricbeat] Fix changelog (#15681) * Fix changelog * ci: use APM pipeline library (#15636) it uses APM pipeline library configured in the instance * AWS Lambda: downgrade Kibana dashboard (#15682) * AWS Lambda: downgrade Kibana dashboard * Downgrade other AWS dashboards * Log command error if setup dashboards fails * Another downgrade * Use github.com/godror/godror instead of goracle.v2 (#15683) From the README of goracle: > Goracle is deprecated because of naming (trademark) issues. From now on we are using github.com/godror/godror instead. * Move pdh query to shared location in order for new modules/metricsets to reuse (#15503) * Move pdh query to shared location * Update changelog * Fix make update * mage fmt * fix changelog * Remove datasource option from SQL module and add tests (#15686) Remove datasource option from SQL module. This option was intended to set the DSN of a database connection, and we were ignoring the hosts setting. In other SQL modules we are using the values in hosts as DSNs, do here the same for consistency. Host is redacted when we cannot parse it as it can contain passwords. StandardizeEvent is exposed in mbtest.Fetcher interface so we can more easily check contents of events in tests. Add integration tests of the module with MySQL and PostgreSQL. Add real data.json with data from MySQL and PostgreSQL. * [metricbeat] add service metricset to reference documentation (#15643) * add service metricset to ref docs * update xpack docs * [metricbeat] Add divide by zero check to docker/diskio (#15649) * add NaN check to docker/diskio * return 0 * Change joda style pattern to java style (#15695) since 7.0 elasticsearch is using java.time style patterns. YYYY becomes yyyy * [DOCS] Add missing config options to shared file (#15136) * [DOCS] Add missing config options to shared file * Add fixes from review * Run mage fmt update to fix build error * [Heartbeat] Support for multiple status codes #13595 (#15587) Allow for multiple status codes in config. Fixes #13595 * Add missing changelog entry for #15587 (#15721) * Update github.com/godror/godror to v0.10.4 (#15737) ## What does this PR do? This PR updates the dependency `github.com/godror/godror` to v0.10.4. ## Why is it important? Packaging of Metricbeat fails due to the issue in the `godror` version we are currently using. See more about the problem: godror/godror#8 * Collect normalized CPU percentages by default (#15729) * Collect normalized CPU percentages by default * Adding CHANGELOG entry * Updating x-pack/metricbeat * Fix: mage check * Detect Eclipse Public License Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com> * [Filebeat] Refactor mqtt input (#16014) * Refactor mqtt input * Fix: comment * Add unit tests * Test: input run * Fix Test: run and stop * Test: backoff * Adjust code after review * MQTT: update docs (#16152) * MQTT: add integration test (#16143) * Create mosquitto image * MQTT input: add integration test * Fix * Verify connectivity * Fix * Fix: mage check * Fix * Fix * Fix: remove global var * Update changelog * Fix: regenerate notice file * Remove unused dependency * Fix: zero qos * Wait asynchronously for client being disconnected Co-authored-by: Felix <felix.roessel@elastic.co> Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com> (cherry picked from commit 56c93e5) * Update CHANGELOG

* Add MQTT input to Filebeat (elastic#15287) * Inital commit for MQTT input * Improved naming and error handling * Improved naming and connection procedure * Merge "master" branch into "feature-mqtt-input" (elastic#15745) * [Filebeat] Fixes for NetFlow v9 devices from various vendors (elastic#15449) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes elastic#14212 * update settings for `decode_csv_fields` (elastic#15249) (elastic#15550) Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> * docs: updates to output config (elastic#15272) * [Filebeat] Handle error message in handleS3Objects function (elastic#15545) * Handle error message in handleS3Objects function * remove s3Context.Fail and use setError and done instead * Add changelog * Fix use of wrong fields in Cisco ASA dashboard (elastic#15553) This dashboard wasn't updated after a couple of fields were renamed. Fixes: elastic#15420 * Add test for publisher spool encode and decode. (elastic#15534) * Add test for publisher queue encode and decode. * Run mage fmt. * Fixes from code review. * Fix convert processor conversion of string with leading zeros to integer (elastic#15557) The conversion failed when for strings with leading zeroes and a decimal digit 8 or 9, as the underlying runtime function would try to parse that as an octal number. This is fixed by only allowing decimal and hex, which in turns makes the processor more aligned to its Elasticsearch counterpart. Fixes elastic#15513 * New mage target: generate pkg file to test the manager (elastic#15580) This PR adds a new mage target to Functionbeat named `buildPkgForFunction`. It generates the folder `pkg` with the functions to make testing the manager more comfortable during development. * Packetbeat TLS: Replace array fields with keyword (elastic#15597) Use of `type: array` in some fields (which was inconsistent) causes those fields to be excluded from the template. This prevents pointing aliases to those fields, which we need in 7.6+. Setting those fields to `keyword` explicitly so that they are included in the template. Fixes elastic#15588 * Add a pull request template providing valuable information when reviewing a PR (elastic#15388) * Add a PR template that provides valuable information when reviewing a PR * Add CLA check * Fix typo * Address comments during review * SF: Fix typo * Add deprecation as PR type * Make it clear how to strike through in markdown * Add default configuration files to the checklist * [Metricbeat] Implement IBM MQ module (elastic#15301) * Modify cockroachdb source * Define testdata * Do not publish ports * Update docs * mage fmt update * Describe containerized environment * Update CHANGELOG.next.asciidoc Co-Authored-By: Chris Mark <chrismarkou92@gmail.com> * Update data.json * Rename image * Update source after review * Filter ibmmq_ metrics * mage check * Fix: mage check * Don't expose port * Rename status to qmgr * Add subscriptions overview dashboard for IBM MQ module * Add calls, messages overview dashboard for IBM MQ module * Add screenshots * Fix: mage check * Fix: CHANGELOG * Add explanation * Fix: mage check Co-authored-by: Chris Mark <chrismarkou92@gmail.com> * Cleanup changelogs for master (elastic#15617) * Cleanup changelogs for master * Remove extra header in CHANGELOG.asciidoc * [Metricbeat] Add lambda metricset in aws module (elastic#15260) * Add lambda metricset * Adds missing imports (elastic#15624) * [docs] Clarify privileges required for the writer role (elastic#15604) * Mask password discovered via module autodiscover hint (elastic#15616) * Mask password is string representation of config * Rename method * Adding unit test * Use const for module config password setting name * Using common.DebugString * Simplifying * Removing now-invalid unit test * Removing now-unnecessary const * Refactoring: moving debug-related var and func to common file * Refactoring: rename from black list to mask list * Implement fmt.Formatter for common.MapStr * Reintroduce debug statement * Make MarshalLogObject always filter MapStr object for logging purposes * Refactoring: renaming to be bit more generic * Forgot to add license header to new file * Fixing verb syntax * Update KQL to get estimated cost without dimension ServiceName (elastic#15640) * Adding monitoring.cloud.* settings to reference files (elastic#15648) * Adding monitoring.cloud.* settings to reference files * Missed winlogbeat somehow * Missed x-pack/winlogbeat * remove lablels (elastic#15644) * Fix panic: don't send events if client is nil (elastic#15568) * Fix panic: don't send events if client is nil * Use mutex * Add CHANGELOG entry * Rename changelog entry * Fix: changelog * Temporarily use specific logstash release * [Metricbeat] Add Istio mesh metricset (elastic#15535) * [Metricbeat] Fix changelog (elastic#15681) * Fix changelog * ci: use APM pipeline library (elastic#15636) it uses APM pipeline library configured in the instance * AWS Lambda: downgrade Kibana dashboard (elastic#15682) * AWS Lambda: downgrade Kibana dashboard * Downgrade other AWS dashboards * Log command error if setup dashboards fails * Another downgrade * Use github.com/godror/godror instead of goracle.v2 (elastic#15683) From the README of goracle: > Goracle is deprecated because of naming (trademark) issues. From now on we are using github.com/godror/godror instead. * Move pdh query to shared location in order for new modules/metricsets to reuse (elastic#15503) * Move pdh query to shared location * Update changelog * Fix make update * mage fmt * fix changelog * Remove datasource option from SQL module and add tests (elastic#15686) Remove datasource option from SQL module. This option was intended to set the DSN of a database connection, and we were ignoring the hosts setting. In other SQL modules we are using the values in hosts as DSNs, do here the same for consistency. Host is redacted when we cannot parse it as it can contain passwords. StandardizeEvent is exposed in mbtest.Fetcher interface so we can more easily check contents of events in tests. Add integration tests of the module with MySQL and PostgreSQL. Add real data.json with data from MySQL and PostgreSQL. * [metricbeat] add service metricset to reference documentation (elastic#15643) * add service metricset to ref docs * update xpack docs * [metricbeat] Add divide by zero check to docker/diskio (elastic#15649) * add NaN check to docker/diskio * return 0 * Change joda style pattern to java style (elastic#15695) since 7.0 elasticsearch is using java.time style patterns. YYYY becomes yyyy * [DOCS] Add missing config options to shared file (elastic#15136) * [DOCS] Add missing config options to shared file * Add fixes from review * Run mage fmt update to fix build error * [Heartbeat] Support for multiple status codes elastic#13595 (elastic#15587) Allow for multiple status codes in config. Fixes elastic#13595 * Add missing changelog entry for elastic#15587 (elastic#15721) * Update github.com/godror/godror to v0.10.4 (elastic#15737) ## What does this PR do? This PR updates the dependency `github.com/godror/godror` to v0.10.4. ## Why is it important? Packaging of Metricbeat fails due to the issue in the `godror` version we are currently using. See more about the problem: godror/godror#8 * Collect normalized CPU percentages by default (elastic#15729) * Collect normalized CPU percentages by default * Adding CHANGELOG entry * Updating x-pack/metricbeat * Fix: mage check * Detect Eclipse Public License Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com> * [Filebeat] Refactor mqtt input (elastic#16014) * Refactor mqtt input * Fix: comment * Add unit tests * Test: input run * Fix Test: run and stop * Test: backoff * Adjust code after review * MQTT: update docs (elastic#16152) * MQTT: add integration test (elastic#16143) * Create mosquitto image * MQTT input: add integration test * Fix * Verify connectivity * Fix * Fix: mage check * Fix * Fix * Fix: remove global var * Update changelog * Fix: regenerate notice file * Remove unused dependency * Fix: zero qos * Wait asynchronously for client being disconnected Co-authored-by: Felix <felix.roessel@elastic.co> Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com>

…#15449) (elastic#15555) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes elastic#14212 (cherry picked from commit 10262cf)

* [Filebeat] Fixes for NetFlow v9 devices from various vendors (elastic#15449) - Allow for zero scope fields in options template NetFlow v9 spec allows for options templates that contain no scope fields. The netflow input was treating this case as an error and discarding the template, but that is only applicable to IPFIX. - Use additional fields to populate bytes/pkt counters Some devices out there (Cisco NSEL) use fields 231/232 as bytes counters, when those are supposed to be layer 4 payload counters. This updates the ECS fields populator to use those fields when the expected ones are not found. - Support a classId of 32 bits While the spec mandates a classId of 8 bits, some Cisco ASA devices actually use a 32 bit version of this field. This patches the field to allow up to 32-bit integers and updates the index pattern to use `long` for the `netflow.class_id` field. - Add more fields from v9 Cisco devices Fixes elastic#14212 * update settings for `decode_csv_fields` (elastic#15249) (elastic#15550) Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> * docs: updates to output config (elastic#15272) * [Filebeat] Handle error message in handleS3Objects function (elastic#15545) * Handle error message in handleS3Objects function * remove s3Context.Fail and use setError and done instead * Add changelog * Fix use of wrong fields in Cisco ASA dashboard (elastic#15553) This dashboard wasn't updated after a couple of fields were renamed. Fixes: elastic#15420 * Add test for publisher spool encode and decode. (elastic#15534) * Add test for publisher queue encode and decode. * Run mage fmt. * Fixes from code review. * Fix convert processor conversion of string with leading zeros to integer (elastic#15557) The conversion failed when for strings with leading zeroes and a decimal digit 8 or 9, as the underlying runtime function would try to parse that as an octal number. This is fixed by only allowing decimal and hex, which in turns makes the processor more aligned to its Elasticsearch counterpart. Fixes elastic#15513 * New mage target: generate pkg file to test the manager (elastic#15580) This PR adds a new mage target to Functionbeat named `buildPkgForFunction`. It generates the folder `pkg` with the functions to make testing the manager more comfortable during development. * Packetbeat TLS: Replace array fields with keyword (elastic#15597) Use of `type: array` in some fields (which was inconsistent) causes those fields to be excluded from the template. This prevents pointing aliases to those fields, which we need in 7.6+. Setting those fields to `keyword` explicitly so that they are included in the template. Fixes elastic#15588 * Add a pull request template providing valuable information when reviewing a PR (elastic#15388) * Add a PR template that provides valuable information when reviewing a PR * Add CLA check * Fix typo * Address comments during review * SF: Fix typo * Add deprecation as PR type * Make it clear how to strike through in markdown * Add default configuration files to the checklist * [Metricbeat] Implement IBM MQ module (elastic#15301) * Modify cockroachdb source * Define testdata * Do not publish ports * Update docs * mage fmt update * Describe containerized environment * Update CHANGELOG.next.asciidoc Co-Authored-By: Chris Mark <chrismarkou92@gmail.com> * Update data.json * Rename image * Update source after review * Filter ibmmq_ metrics * mage check * Fix: mage check * Don't expose port * Rename status to qmgr * Add subscriptions overview dashboard for IBM MQ module * Add calls, messages overview dashboard for IBM MQ module * Add screenshots * Fix: mage check * Fix: CHANGELOG * Add explanation * Fix: mage check Co-authored-by: Chris Mark <chrismarkou92@gmail.com> * Cleanup changelogs for master (elastic#15617) * Cleanup changelogs for master * Remove extra header in CHANGELOG.asciidoc * [Metricbeat] Add lambda metricset in aws module (elastic#15260) * Add lambda metricset * Adds missing imports (elastic#15624) * [docs] Clarify privileges required for the writer role (elastic#15604) * Mask password discovered via module autodiscover hint (elastic#15616) * Mask password is string representation of config * Rename method * Adding unit test * Use const for module config password setting name * Using common.DebugString * Simplifying * Removing now-invalid unit test * Removing now-unnecessary const * Refactoring: moving debug-related var and func to common file * Refactoring: rename from black list to mask list * Implement fmt.Formatter for common.MapStr * Reintroduce debug statement * Make MarshalLogObject always filter MapStr object for logging purposes * Refactoring: renaming to be bit more generic * Forgot to add license header to new file * Fixing verb syntax * Update KQL to get estimated cost without dimension ServiceName (elastic#15640) * Adding monitoring.cloud.* settings to reference files (elastic#15648) * Adding monitoring.cloud.* settings to reference files * Missed winlogbeat somehow * Missed x-pack/winlogbeat * remove lablels (elastic#15644) * Fix panic: don't send events if client is nil (elastic#15568) * Fix panic: don't send events if client is nil * Use mutex * Add CHANGELOG entry * Rename changelog entry * Fix: changelog * Temporarily use specific logstash release * [Metricbeat] Add Istio mesh metricset (elastic#15535) * [Metricbeat] Fix changelog (elastic#15681) * Fix changelog * ci: use APM pipeline library (elastic#15636) it uses APM pipeline library configured in the instance * AWS Lambda: downgrade Kibana dashboard (elastic#15682) * AWS Lambda: downgrade Kibana dashboard * Downgrade other AWS dashboards * Log command error if setup dashboards fails * Another downgrade * Use github.com/godror/godror instead of goracle.v2 (elastic#15683) From the README of goracle: > Goracle is deprecated because of naming (trademark) issues. From now on we are using github.com/godror/godror instead. * Move pdh query to shared location in order for new modules/metricsets to reuse (elastic#15503) * Move pdh query to shared location * Update changelog * Fix make update * mage fmt * fix changelog * Remove datasource option from SQL module and add tests (elastic#15686) Remove datasource option from SQL module. This option was intended to set the DSN of a database connection, and we were ignoring the hosts setting. In other SQL modules we are using the values in hosts as DSNs, do here the same for consistency. Host is redacted when we cannot parse it as it can contain passwords. StandardizeEvent is exposed in mbtest.Fetcher interface so we can more easily check contents of events in tests. Add integration tests of the module with MySQL and PostgreSQL. Add real data.json with data from MySQL and PostgreSQL. * [metricbeat] add service metricset to reference documentation (elastic#15643) * add service metricset to ref docs * update xpack docs * [metricbeat] Add divide by zero check to docker/diskio (elastic#15649) * add NaN check to docker/diskio * return 0 * Change joda style pattern to java style (elastic#15695) since 7.0 elasticsearch is using java.time style patterns. YYYY becomes yyyy * [DOCS] Add missing config options to shared file (elastic#15136) * [DOCS] Add missing config options to shared file * Add fixes from review * Run mage fmt update to fix build error * [Heartbeat] Support for multiple status codes elastic#13595 (elastic#15587) Allow for multiple status codes in config. Fixes elastic#13595 * Add missing changelog entry for elastic#15587 (elastic#15721) * Update github.com/godror/godror to v0.10.4 (elastic#15737) ## What does this PR do? This PR updates the dependency `github.com/godror/godror` to v0.10.4. ## Why is it important? Packaging of Metricbeat fails due to the issue in the `godror` version we are currently using. See more about the problem: godror/godror#8 * Collect normalized CPU percentages by default (elastic#15729) * Collect normalized CPU percentages by default * Adding CHANGELOG entry * Updating x-pack/metricbeat * Fix: mage check * Detect Eclipse Public License Co-authored-by: Adrian Serrano <adrisr83@gmail.com> Co-authored-by: DeDe Morton <dede.morton@elastic.co> Co-authored-by: Sophia Xu <sophia.xu@elastic.co> Co-authored-by: Brandon Morelli <bmorelli25@gmail.com> Co-authored-by: kaiyan-sheng <kaiyan.sheng@elastic.co> Co-authored-by: Blake Rouse <blake.rouse@elastic.co> Co-authored-by: Noémi Ványi <kvch@users.noreply.github.com> Co-authored-by: Manuel de la Peña <social.mdelapenya@gmail.com> Co-authored-by: Chris Mark <chrismarkou92@gmail.com> Co-authored-by: Michael Madden <mikemadden42@users.noreply.github.com> Co-authored-by: Shaunak Kashyap <ycombinator@gmail.com> Co-authored-by: Pablo Mercado <pablo.mercado@elastic.co> Co-authored-by: Ivan Fernandez Calvo <kuisathaverat@users.noreply.github.com> Co-authored-by: Mariana Dima <mariana@elastic.co> Co-authored-by: Jaime Soriano Pastor <jaime.soriano@elastic.co> Co-authored-by: Alex K. <8418476+fearful-symmetry@users.noreply.github.com> Co-authored-by: Przemyslaw Gomulka <przemyslaw.gomulka@elastic.co> Co-authored-by: Amanda H. L. de Andrade Katz <amanda.andrade@serpro.gov.br> Co-authored-by: Andrew Cholakian <andrew@andrewvc.com>

adriansr transferred this issue from elastic/kibana Oct 23, 2019

andrewkroh closed this as completed Dec 10, 2019

adriansr self-assigned this Jan 7, 2020

adriansr reopened this Jan 7, 2020

adriansr mentioned this issue Jan 14, 2020

[Filebeat] Fixes for NetFlow v9 devices from various vendors #15449

Merged

4 tasks

adriansr closed this as completed in #15449 Jan 14, 2020

adriansr mentioned this issue Jan 14, 2020

Cherry-pick #15449 to 7.x: [Filebeat] Fixes for NetFlow v9 devices from various vendors #15554

Merged

4 tasks

adriansr mentioned this issue Jan 14, 2020

Cherry-pick #15449 to 7.5: [Filebeat] Fixes for NetFlow v9 devices from various vendors #15555

Merged

4 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[SIEM] Netflow Data always displays 0 bytes in / out #14212

[SIEM] Netflow Data always displays 0 bytes in / out #14212

StefanSa commented Oct 12, 2019

elasticmachine commented Oct 15, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019

kicker83 commented Oct 22, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019 •

edited

Loading

kicker83 commented Oct 23, 2019 via email •

edited

Loading

adriansr commented Oct 23, 2019

StefanSa commented Dec 3, 2019

StefanSa commented Dec 6, 2019

andrewkroh commented Dec 10, 2019

adriansr commented Dec 10, 2019

StefanSa commented Dec 10, 2019

adriansr commented Jan 7, 2020 •

edited

Loading

[SIEM] Netflow Data always displays 0 bytes in / out #14212

[SIEM] Netflow Data always displays 0 bytes in / out #14212

Comments

StefanSa commented Oct 12, 2019

elasticmachine commented Oct 15, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019

kicker83 commented Oct 22, 2019

kicker83 commented Oct 22, 2019

adriansr commented Oct 22, 2019 • edited Loading

kicker83 commented Oct 23, 2019 via email • edited Loading

adriansr commented Oct 23, 2019

StefanSa commented Dec 3, 2019

StefanSa commented Dec 6, 2019

andrewkroh commented Dec 10, 2019

adriansr commented Dec 10, 2019

StefanSa commented Dec 10, 2019

adriansr commented Jan 7, 2020 • edited Loading

adriansr commented Oct 22, 2019 •

edited

Loading

kicker83 commented Oct 23, 2019 via email •

edited

Loading

adriansr commented Jan 7, 2020 •

edited

Loading