|
Mantis - Resin
|
|||||
| Viewing Issue Advanced Details | |||||
|
|
|||||
| ID: | Category: | Severity: | Reproducibility: | Date Submitted: | Last Update: |
| 10 | minor | always | 11-29-05 10:31 | 03-30-06 09:17 | |
|
|
|||||
| Reporter: | ferg | Platform: | |||
| Assigned To: | ferg | OS: | |||
| Priority: | normal | OS Version: | |||
| Status: | closed | Product Version: | |||
| Product Build: | Resolution: | fixed | |||
| Projection: | none | ||||
| ETA: | none | Fixed in Version: | 3.0.17 | ||
|
|
|||||
| Summary: | 0000010: 3.0.14 startup delays with cluster | ||||
| Description: |
(rep by Galen O'Hanlon) With our current configuration running Resin 3.0.14, when any node of the cluster is starting up the entire cluster is unable to service requests. We run many web applications and several of the applications take a long time to start which will result in long periods of downtime should any single node fail. We have also confirmed that Resin 3.0.15 has the same behavior. The test cases described below both work as expected under Resin 3.0.12. That is, the nodes do not appear to affect each other at all during startup. Additionally, we observed that the message “hmux listening to prodc01:6804” appears before the hosts begin starting in the 3.0.14 output, but appears after all the hosts have finished starting in 3.0.12. Attachments: front.conf.xml: Front-end Resin configuration file. back.conf.xml: Back-end Resin configuration file. foo.war: A simple web application with a delay of configurable duration placed in the “init” method. It is currently configured to be a 90 second delay, long enough to cause a timeout exception during Resin startup on a sibling node. (see mail for attachments) |
||||
| Steps To Reproduce: | |||||
| Additional Information: | |||||
| Relationships | |||||
| Attached Files: | |||||
| There are no notes attached to this issue. |