Back to bug 1717011
| Who | When | What | Removed | Added |
|---|---|---|---|---|
| Manjunatha | 2019-06-04 14:26:55 UTC | Priority | unspecified | medium |
| Manjunatha | 2019-06-04 14:27:15 UTC | Severity | unspecified | medium |
| Ashish Singh | 2019-06-06 06:40:12 UTC | CC | assingh | |
| Dimitri Savineau | 2019-06-06 18:23:44 UTC | Status | NEW | ASSIGNED |
| CC | dsavinea | |||
| Link ID | Github ceph/ceph-ansible/pull/4059 | |||
| Assignee | gabrioux | dsavinea | ||
| Flags | needinfo?(mmanjuna) | |||
| Manjunatha | 2019-06-07 09:49:06 UTC | Flags | needinfo?(mmanjuna) | |
| Dimitri Savineau | 2019-06-12 21:02:03 UTC | Link ID | Github ceph/ceph-ansible/pull/4095 Github ceph/ceph-ansible/pull/4094 | |
| Guillaume Abrioux | 2019-06-13 16:55:04 UTC | Status | ASSIGNED | POST |
| CC | gabrioux | |||
| Tejas | 2019-06-17 05:13:54 UTC | Status | POST | MODIFIED |
| CC | tserlin | |||
| Fixed In Version | ceph-ansible-4.0.0-0.1.rc9.el8cp | |||
| CC | ceph-qe-bugs | |||
| Flags | needinfo?(ceph-qe-bugs) | |||
| CC | tchandra | |||
| Flags | needinfo?(ceph-qe-bugs) | |||
| errata-xmlrpc | 2019-06-17 14:46:37 UTC | Status | MODIFIED | ON_QA |
| Giridhar Ramaraju | 2019-08-05 13:10:53 UTC | QA Contact | ceph-qe-bugs | hgurav |
| Hemant G | 2019-08-06 08:36:16 UTC | QA Contact | hgurav | vashastr |
| PnT Account Manager | 2019-09-02 10:06:01 UTC | CC | sankarshan | |
| Bara Ancincova | 2019-12-10 10:36:13 UTC | Blocks | 1730176 | |
| Doc Type | If docs needed, set a value | Bug Fix | ||
| Dimitri Savineau | 2020-01-07 21:10:48 UTC | Doc Text | Cause: On a system running thousands processes, the fuser command could take several minutes or hours to complete because it iterates over all PID present in /proc Consequence: The handler tasks checking if a ceph process is already running takes a long time to complete. Fix: Instead of using the fuser command we can check the socket file in /proc/net/unix. Result: The handler tasks checking the ceph socket is almost instant to complete. |
|
| Erin Donnelly | 2020-01-08 16:51:19 UTC | CC | edonnell | |
| Doc Text | Cause: On a system running thousands processes, the fuser command could take several minutes or hours to complete because it iterates over all PID present in /proc Consequence: The handler tasks checking if a ceph process is already running takes a long time to complete. Fix: Instead of using the fuser command we can check the socket file in /proc/net/unix. Result: The handler tasks checking the ceph socket is almost instant to complete. | .The Ansible playbook no longer takes hours to complete the `fuser` command Previously, on a system running thousands of processes, the `fuser` command in the Ansible playbook could take several minutes or hours to complete because it iterates over all PID present in the `/proc` directory. Because of this, the handler tasks take a long time to complete checking if a Ceph process is already running. With this update, instead of using the `fuser` command, the Ansible playbook now checks the socket file in the `/proc/net/unix` directory, and the handler tasks that check the Ceph socket completes almost instantly. | ||
| Flags | needinfo?(dsavinea) | |||
| Dimitri Savineau | 2020-01-08 18:22:46 UTC | Flags | needinfo?(dsavinea) | |
| Bara Ancincova | 2020-01-13 11:37:27 UTC | Docs Contact | edonnell | |
| errata-xmlrpc | 2020-01-31 11:27:51 UTC | Status | ON_QA | RELEASE_PENDING |
| errata-xmlrpc | 2020-01-31 12:46:18 UTC | Status | RELEASE_PENDING | CLOSED |
| Resolution | --- | ERRATA | ||
| Last Closed | 2020-01-31 12:46:18 UTC | |||
| errata-xmlrpc | 2020-01-31 12:46:42 UTC | Link ID | Red Hat Product Errata RHBA-2020:0312 |
Back to bug 1717011