Petals ESB Container

Too many connections established

Details

  • Type: Bug Bug
  • Status: New New
  • Priority: Blocker Blocker
  • Resolution: Unresolved
  • Affects Version/s: 5.0.0
  • Fix Version/s: 5.4.0
  • Component/s: Transporter
  • Security Level: Public
  • Description:
    Hide

    During load tests in the cloud with Roboconf we can see that too many connections are established by the NIO transporter.

    Number of connections of the NIO transporter accepted by container-bootstrap from (Petals ESB containers are not restarted between 2 runs):

    Container Nb connection established during the 1st run Nb connection established during the 2nd run
    container-1 317 15
    autonomic-0 105 9
    autonomic-1 35 9
    autonomic-2 10 9
    autonomic-3
    		10

    After the 2nd run, the number of connections coming from container-bootstrap into:

    Container After the 2nd run
    container-1 1694
    autonomic-0 965
    autonomic-1 544
    autonomic-2 341
    autonomic-3 17

    I think there are a lot of connection established, and this has impact on performances

    Show
    During load tests in the cloud with Roboconf we can see that too many connections are established by the NIO transporter. Number of connections of the NIO transporter accepted by container-bootstrap from (Petals ESB containers are not restarted between 2 runs):
    Container Nb connection established during the 1st run Nb connection established during the 2nd run
    container-1 317 15
    autonomic-0 105 9
    autonomic-1 35 9
    autonomic-2 10 9
    autonomic-3
    		10
    After the 2nd run, the number of connections coming from container-bootstrap into:
    Container After the 2nd run
    container-1 1694
    autonomic-0 965
    autonomic-1 544
    autonomic-2 341
    autonomic-3 17
    I think there are a lot of connection established, and this has impact on performances
  • Environment:
    -

Issue Links

Depends

Activity

Descending order - Click to sort in ascending order
Hide
Permalink
Christophe DENEUX added a comment - Thu, 13 Apr 2023 - 17:02:55 +0200

Postponed to version 5.4.0 or upper

Show
Christophe DENEUX added a comment - Thu, 13 Apr 2023 - 17:02:55 +0200 Postponed to version 5.4.0 or upper
Hide
Permalink
Christophe DENEUX added a comment - Mon, 19 Sep 2016 - 15:00:39 +0200

This issue should be solved with PETALSESBCONT-437 (Reimplement the Remote TCP transporter with Apache Netty)

Show
Christophe DENEUX added a comment - Mon, 19 Sep 2016 - 15:00:39 +0200 This issue should be solved with PETALSESBCONT-437 (Reimplement the Remote TCP transporter with Apache Netty)
Hide
Permalink
Christophe DENEUX added a comment - Mon, 15 Feb 2016 - 18:06:16 +0100 - edited

The metrics about connections show that a lot of connections are established without any reason:

petals-cli@172.17.0.7:7700>monitoring -o tcp-transporter -f outgoing-connections
| RemoteContainer          || ActiveConnectionsCurrent | ActiveConnectionsMax | IdleConnectionsCurrent | IdleConnectionsMax | MinSize | MaxSize | EstablishedConnections |
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| container-bootstrap-node || 0                        | 0                    | 0                      | 0                  | 0       |      10 |                      0 |
| container-cons-node-1    || 0                        | 0                    | 0                      | 0                  | 0       |      10 |                      0 |
| container-prov-node-2    || 0                        | 10                   | 0                      | 10                 | 0       |      10 |                   2545 |
| container-prov-node-1    || 0                        | 10                   | 1                      | 10                 | 0       |      10 |                   2448 |
Show
Christophe DENEUX added a comment - Mon, 15 Feb 2016 - 18:06:16 +0100 - edited The metrics about connections show that a lot of connections are established without any reason: 
petals-cli@172.17.0.7:7700>monitoring -o tcp-transporter -f outgoing-connections
| RemoteContainer          || ActiveConnectionsCurrent | ActiveConnectionsMax | IdleConnectionsCurrent | IdleConnectionsMax | MinSize | MaxSize | EstablishedConnections |
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| container-bootstrap-node || 0                        | 0                    | 0                      | 0                  | 0       |      10 |                      0 |
| container-cons-node-1    || 0                        | 0                    | 0                      | 0                  | 0       |      10 |                      0 |
| container-prov-node-2    || 0                        | 10                   | 0                      | 10                 | 0       |      10 |                   2545 |
| container-prov-node-1    || 0                        | 10                   | 1                      | 10                 | 0       |      10 |                   2448 |
Hide
Permalink
Christophe DENEUX added a comment - Wed, 10 Feb 2016 - 14:54:46 +0100

To improve the statistics mentioned (that are based on content of log files), we can add metrics about connections of the NIO transporter at container level: PETALSESBCONT-395

Show
Christophe DENEUX added a comment - Wed, 10 Feb 2016 - 14:54:46 +0100 To improve the statistics mentioned (that are based on content of log files), we can add metrics about connections of the NIO transporter at container level: PETALSESBCONT-395

People

Dates

  • Created:
    Thu, 29 Oct 2015 - 12:39:44 +0100
    Updated:
    Thu, 13 Apr 2023 - 17:02:55 +0200
Atlassian JIRA (v4.1.2#531) | Report a problem | Request a feature | Contact administrators
Powered by a free Atlassian JIRA open source license for Petals ESB. Try JIRA - bug tracking software for your team.