Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Can’t parse haproxy logs #25827

Closed
manuelsaks opened this issue May 24, 2021 · 19 comments · Fixed by #25835
Closed

Can’t parse haproxy logs #25827

manuelsaks opened this issue May 24, 2021 · 19 comments · Fixed by #25835
Labels
Team:Integrations Label for the Integrations team

Comments

@manuelsaks
Copy link

I created Grok pattern which works in Kibana Debug Grok devtool:
Sample data:

[May 22 02:22:22 server1 haproxy[5089]: -:22222 [22/May/2021:02:22:22.222] www-https~ myapp/node2 site.domain.com 0/0/0/18/18 200 200 - - ---- 222/222/2/0/0 0/0 \"OPTIONS /api/v2/app/ HTTP/1.1\"]

Grok pattern:

%{HAPROXY_LOG_DATE:haproxy.logdate} %{NOTSPACE:haproxy.host} %{NOTSPACE:process.name[pid]}[%{NUMBER:process.pid:long}] (%{IP:source.address}|-):%{POSINT:source} %{HAPROXY_DATE:haproxy.request_date} %{NOTSPACE:haproxy.frontend_name} %{NOTSPACE:haproxy.backend_name}/%{NOTSPACE:haproxy.server_name} %{NOTSPACE:haproxy.http.captured.request.headers} %{NUMBER:haproxy.http.request.time_wait_ms:long}/%{NUMBER:haproxy.total_waiting_time_ms:long}/%{NUMBER:haproxy.connection_wait_time_ms:long}/%{NUMBER:haproxy.http.request.time_wait_without_data_ms:long}/%{NUMBER:temp.duration:long} %{NUMBER:http.response.status_code:long} %{NUMBER:haproxy.bytes_read:long} %{NOTSPACE:haproxy.http.request.captured_cookie} %{NOTSPACE:haproxy.http.response.captured_cookie} %{NOTSPACE:haproxy.termination_state} %{NUMBER:haproxy.connections.active:long}/%{NUMBER:haproxy.connections.frontend:long}/%{NUMBER:haproxy.connections.backend:long}/%{NUMBER:haproxy.connections.server:long}/%{NUMBER:haproxy.connections.retries:long} %{NUMBER:haproxy.server_queue:long}/%{NUMBER:haproxy.backend_queue:long} \\\"%{NOTSPACE:haproxy.http.request.method} %{NOTSPACE:haproxy.http.request.captured_headers} %{NOTSPACE:haproxy.http.response.captured_headers}\\\"

Custom patterns:

HAPROXY_LOG_DATE %{MONTH} %{MONTHDAY} %{HOUR}:%{MINUTE}:%{SECOND}
HAPROXY_DATE \[%{MONTHDAY}[/-]%{MONTH}[/-]%{YEAR}:%{HOUR}:%{MINUTE}:%{SECOND}\]

Structured data:

{
  "process": {
    "name[pid]": "haproxy[5089]"
  },
  "temp": {
    "duration": 18
  },
  "haproxy": {
    "server_name": "node2",
    "total_waiting_time_ms": 0,
    "termination_state": "----",
    "connection_wait_time_ms": 0,
    "bytes_read": 200,
    "backend_queue": 0,
    "backend_name": "myapp",
    "logdate": "May 22 02:22:22",
    "host": "server1",
    "request_date": "[22/May/2021:02:22:22.222]",
    "http": {
      "request": {
        "captured_cookie": "-",
        "time_wait_without_data_ms": 18,
        "captured_headers": "/api/v2/app/",
        "method": "OPTIONS",
        "time_wait_ms": 0
      },
      "response": {
        "captured_cookie": "-",
        "captured_headers": "HTTP/1.1"
      },
      "captured": {
        "request": {
          "headers": "site.domain.com"
        }
      }
    },
    "frontend_name": "www-https~",
    "server_queue": 0,
    "connections": {
      "server": 0,
      "retries": 0,
      "active": 222,
      "backend": 2,
      "frontend": 222
    }
  },
  "http": {
    "response": {
      "status_code": 200
    }
  },
  "source": "22222"
}

First I thought that something else causes the problem: discuss.elastic.co

But the pattern is working, and I don't know why. but it doesn't work with Filebeat and Elastic.
I have an error in Elastic:

Provided Grok expressions do not match field value:
[May 22 02:22:22 server1 haproxy[5089]: -:22222 [22/May/2021:02:22:22.222] www-https~ myapp/node2 site.domain.com 0/0/0/18/18 200 200 - - ---- 222/222/2/0/0 0/0 \"OPTIONS /api/v2/app/ HTTP/1.1\"]

That's my current config file: gist.github.com

  • Version: Filebeat 7.6.1; Elastic&Kibana 7.7.1
  • Operating System: Debian Buster
  • Discuss Forum URL: discuss.elastic.co
@botelastic botelastic bot added the needs_team Indicates that the issue/PR needs a Team:* label label May 24, 2021
@manuelsaks
Copy link
Author

I have the same error. I tried also to upgrade filebeat to 7.12.1 version and copy your pipline.json.
If you have some questions about my issue feel free to ask.

@legoguy1000
Copy link
Contributor

how did you test the changes? I would install the latest filebeat and then update that pipeline with the changes above. I added the sample log that you provided in the ticket. If there are other logs that don't work let me know.

@legoguy1000
Copy link
Contributor

I have the same error. I tried also to upgrade filebeat to 7.12.1 version and copy your pipline.json.
If you have some questions about my issue feel free to ask.

@maarsaks can you provide more details on what didn't work. I added ur sample log to the tests and it parsed properly.

@manuelsaks
Copy link
Author

manuelsaks commented Jun 1, 2021

how did you test the changes? I would install the latest filebeat and then update that pipeline with the changes above. I added the sample log that you provided in the ticket. If there are other logs that don't work let me know.

I've tested on docker images with docker-compose using the latest, fresh version of filebeat.

I have the same error. I tried also to upgrade filebeat to 7.12.1 version and copy your pipline.json.
If you have some questions about my issue feel free to ask.

@maarsaks can you provide more details on what didn't work. I added ur sample log to the tests and it parsed properly.

I'm using Kibana to filter logs and there's is an error: 'Provided Grok expressions do not match field value'.
What details exactly do you need?

@legoguy1000
Copy link
Contributor

How did you update the pipeline? Is the log thats having the issue the same that I added to the sample data??

@manuelsaks
Copy link
Author

manuelsaks commented Jun 1, 2021

How did you update the pipeline?

I updated pipeline by replacing default config with yours in a new, freshly installed filebeat under /usr/share/filebeat/module/haproxy/log/pipline.json.

Is the log thats having the issue the same that I added to the sample data??

Yes.

@legoguy1000
Copy link
Contributor

I updated pipeline by replacing default config with yours in a new, freshly installed filebeat under /usr/share/filebeat/module/haproxy/log/pipline.json.
Ok, and the pipeline loaded into Elasticsearch matches? I've never modified a pipeline that way. I would have done it through Kibana dev tools/API after it was loaded. Can you post the resulting event from ES?

@manuelsaks
Copy link
Author

I updated pipeline by replacing default config with yours in a new, freshly installed filebeat under /usr/share/filebeat/module/haproxy/log/pipline.json.
Ok, and the pipeline loaded into Elasticsearch matches? I've never modified a pipeline that way. I would have done it through Kibana dev tools/API after it was loaded. Can you post the resulting event from ES?

Yup, it loaded and this is the message from ES:

Provided Grok expressions do not match field value:
[May 22 02:22:22 server1 haproxy[5089]: -:22222 [22/May/2021:02:22:22.222] www-https~ myapp/node2 site.domain.com 0/0/0/18/18 200 200 - - ---- 222/222/2/0/0 0/0 \"OPTIONS /api/v2/app/ HTTP/1.1\"]

@legoguy1000
Copy link
Contributor

Hmm, IDK. Here, https://github.com/elastic/beats/pull/25835/files#diff-8363329fdb5b538e99f1112198aaa68d0a353c161a49f26a1041f6f49ae47cfdR63, you can see that the sample log is the same as u provided and is being parsed properly.

@manuelsaks
Copy link
Author

Hmm, IDK. Here, https://github.com/elastic/beats/pull/25835/files#diff-8363329fdb5b538e99f1112198aaa68d0a353c161a49f26a1041f6f49ae47cfdR63, you can see that the sample log is the same as u provided and is being parsed properly.

And that's the problem. Like I said in my first comment the pattern is working in Grok debugger, but it doesn't display correctly in Kibana.

@manuelsaks
Copy link
Author

manuelsaks commented Jun 9, 2021

@legoguy1000

I've created the repo how to reproduce the error in ~5 minutes - make sure that you've made the all steps from README.md.
github.com/maarsaks/elk-docker-compose

@legoguy1000
Copy link
Contributor

@legoguy1000

I've created the repo how to reproduce the error in ~5 minutes - make sure that you've made the all steps from README.md.
github.com/maarsaks/elk-docker-compose

I'm looking at your pipeline file, you file doesn't have any of the changes that I made in the PR. What are you basing that file off of? This is the file you should copy, https://github.com/elastic/beats/blob/5582e6ccfddd152e4bcf4141361d6a0e0b1e4daa/filebeat/module/haproxy/log/ingest/pipeline.yml.

@manuelsaks
Copy link
Author

Okay, I've updated the repo with the pipline.json that you provided and I have the same issue (from the repo that I've created for reproducing).

@legoguy1000
Copy link
Contributor

Okay, I've updated the repo with the pipline.json that you provided and I have the same issue (from the repo that I've created for reproducing).

So looking at the repo u created I found some issues.

  1. /usr/share/filebeat/module/haproxy/log/ingest/pipline.json: you misspelled pipeline.
  2. You have YAML in a json file
  3. The file you want to overwrite is /usr/share/filebeat/module/haproxy/log/ingest/pipeline.yml

As such your test isn't actually loading the new pipeline. Please validate that the correct pipeline is loaded by going to the ingest pipelines page under stack management in Kibana and look at the HAProxy pipeline. The way I would test this change is load ES, Kibana, and Filebeat normally without any modifications. Manually update the pipeline using the API in dev tools, validate the changes using the API or ingest pipeline page and then ingest data.

@legoguy1000
Copy link
Contributor

@maarsaks we're u able to retest again? If we don't hear anything we're going to go ahead and merge the changes as they pass all the tests with the log samples you provided.

@manuelsaks
Copy link
Author

manuelsaks commented Jun 22, 2021

Thank you.
Those logs that I provided are working.

But there's more :/
These aren't working for example:

Jun 22 10:13:15 node2 haproxy[23034]: [WARNING] 172/101315 (23551) : Server app/node3 is going DOWN for maintenance (entry removed from SRV record). 1 active and 0 backup servers left. 0 sessions active, 0 requeued, 0 remaining in queue.
Jun 22 10:13:15 node2 haproxy[23034]: Server app/node3   is going DOWN for maintenance (entry removed from SRV record). 1 active and 0 backup servers left. 0 sessions active, 0 requeued, 0 remaining in queue.
Jun 22 10:13:15 node2 haproxy[23034]: Server app/node3 is going DOWN for maintenance (entry removed from SRV record). 0 active and 0 backup servers left. 0 sessions active, 0 requeued, 0 remaining in queue.
Jun 22 10:13:13 node2 haproxy[23034]: node/node3 changed its FQDN from (null) to my.domain.name by 'SRV record'
Jun 22 10:13:13 node2 haproxy[23034]: [WARNING] 122/111313 (22553) : app/node3 changed its IP from  to 192.122.0.22 by app/app.
Jun 22 10:13:13 node2 haproxy[23034]: [WARNING] 122/111313 (22553) : Server app/node3 ('node3.domain.eu.com') is UP/READY (resolves again).
Jun 22 10:13:13 node2 haproxy[23034]: Server app/node3 administratively READY thanks to valid DNS answer.
Jun 22 10:13:13 node2 haproxy[23034]: Server app/node4 ('node3.domain.eu.com') is UP/READY (resolves again).
Jun 22 12:02:53 node2 haproxy[23034]: -:47625 [22/Jun/2021:12:02:53.473] www-https~ app/app-node2 app.domain.com 0/0/1/17/18 302 291 - - ---- 1/1/0/0/0 0/0 "GET / HTTP/1.1"
Jun 22 12:03:01 node2 haproxy[23034]: -:47445 [22/Jun/2021:12:03:01.501] www-https~ app/node16 app.domain.com 0/0/1/55/56 200 3097 - - ---- 2/2/0/0/0 0/0 "GET /app/login/ HTTP/1.1"
Jun 22 12:03:01 node2 haproxy[23034]: -:43662 [22/Jun/2021:12:03:01.427] www-https~ app/node7 app.domain.com 0/0/1/30/31 200 1235 - - ---- 1/1/0/0/0 0/0 "GET /23rfsa/ HTTP/1.1"
Jun 22 12:02:59 node2 haproxy[23034]: -:47481 [22/Jun/2021:12:02:59.590] www-https~ app/node16 app.domain.com 0/0/3/32/35 403 142 - - ---- 1/1/0/0/0 0/0 "GET /app/event/ HTTP/1.1"
Jun 22 12:02:57 node2 haproxy[23034]: -:47642 [22/Jun/2021:12:02:55.202] www-https~ app/app-node2 app.domain.com 1/0/1/15/2606 200 325791 - - ---- 1/1/0/0/0 0/0 "GET /static/files/3rsdfas3.js HTTP/1.1"
Jun 22 12:03:08 node2 haproxy[23034]: -:11178 [22/Jun/2021:12:03:08.833] www-https~ app/node7 app.domain.com 0/0/1/29/30 404 448 - - ---- 3/3/0/0/0 0/0 "GET /favicon.ico HTTP/1.1"
Jun 22 12:03:04 node2 haproxy[23034]: -:21278 [22/Jun/2021:12:03:04.060] www-https~ app/node16 app.domain.com 0/0/2/39/41 200 1235 - - ---- 3/3/0/0/0 0/0 "GET /qfe32/ HTTP/1.1"
Jun 22 12:03:08 node3 haproxy[23034]: -:21978 [22/Jun/2021:12:03:08.339] www-https~ app/server app.domain.eu 0/0/2/45/47 404 448 - - ---- 3/3/0/0/0 0/0 "GET /dsffdssdf HTTP/1.1"

@legoguy1000
Copy link
Contributor

So the module isn't setup to parse the first half of those logs, the server/status messages. As for the 2nd half, they look like they're the same as what I just added so I can take a quick look.

@legoguy1000
Copy link
Contributor

legoguy1000 commented Jun 23, 2021

Jun 22 12:02:53 node2 haproxy[23034]: -:47625 [22/Jun/2021:12:02:53.473] www-https~ app/app-node2 app.domain.com 0/0/1/17/18 302 291 - - ---- 1/1/0/0/0 0/0 "GET / HTTP/1.1"
Jun 22 12:03:01 node2 haproxy[23034]: -:47445 [22/Jun/2021:12:03:01.501] www-https~ app/node16 app.domain.com 0/0/1/55/56 200 3097 - - ---- 2/2/0/0/0 0/0 "GET /app/login/ HTTP/1.1"
Jun 22 12:03:01 node2 haproxy[23034]: -:43662 [22/Jun/2021:12:03:01.427] www-https~ app/node7 app.domain.com 0/0/1/30/31 200 1235 - - ---- 1/1/0/0/0 0/0 "GET /23rfsa/ HTTP/1.1"
Jun 22 12:02:59 node2 haproxy[23034]: -:47481 [22/Jun/2021:12:02:59.590] www-https~ app/node16 app.domain.com 0/0/3/32/35 403 142 - - ---- 1/1/0/0/0 0/0 "GET /app/event/ HTTP/1.1"
Jun 22 12:02:57 node2 haproxy[23034]: -:47642 [22/Jun/2021:12:02:55.202] www-https~ app/app-node2 app.domain.com 1/0/1/15/2606 200 325791 - - ---- 1/1/0/0/0 0/0 "GET /static/files/3rsdfas3.js HTTP/1.1"
Jun 22 12:03:08 node2 haproxy[23034]: -:11178 [22/Jun/2021:12:03:08.833] www-https~ app/node7 app.domain.com 0/0/1/29/30 404 448 - - ---- 3/3/0/0/0 0/0 "GET /favicon.ico HTTP/1.1"
Jun 22 12:03:04 node2 haproxy[23034]: -:21278 [22/Jun/2021:12:03:04.060] www-https~ app/node16 app.domain.com 0/0/2/39/41 200 1235 - - ---- 3/3/0/0/0 0/0 "GET /qfe32/ HTTP/1.1"
Jun 22 12:03:08 node3 haproxy[23034]: -:21978 [22/Jun/2021:12:03:08.339] www-https~ app/server app.domain.eu 0/0/2/45/47 404 448 - - ---- 3/3/0/0/0 0/0 "GET /dsffdssdf HTTP/1.1"

These logs parsed no problem. I think a separate discussion should be have for the other logs. I'm going to recommend that we merge the PR.

legoguy1000 added a commit to legoguy1000/beats that referenced this issue Jun 23, 2021
@jsoriano jsoriano added the Team:Integrations Label for the Integrations team label Jun 23, 2021
@elasticmachine
Copy link
Collaborator

Pinging @elastic/integrations (Team:Integrations)

@botelastic botelastic bot removed the needs_team Indicates that the issue/PR needs a Team:* label label Jun 23, 2021
P1llus pushed a commit that referenced this issue Jun 23, 2021
* #25827: Update HA Proxy log grok patterns

* update changelog

* add more sample data
mergify bot pushed a commit that referenced this issue Jun 23, 2021
* #25827: Update HA Proxy log grok patterns

* update changelog

* add more sample data

(cherry picked from commit a300f1b)
jsoriano pushed a commit that referenced this issue Jun 24, 2021
* #25827: Update HA Proxy log grok patterns

* update changelog

* add more sample data

(cherry picked from commit a300f1b)

Co-authored-by: Alex Resnick <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Team:Integrations Label for the Integrations team
Projects
None yet
Development

Successfully merging a pull request may close this issue.

4 participants