Bug 189381
| Summary: | ATIX - Add nodeid as an option for context dependent path names | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Retired] Red Hat Cluster Suite | Reporter: | Kiersten (Kerri) Anderson <kanderso> | ||||
| Component: | gfs | Assignee: | Abhijith Das <adas> | ||||
| Status: | CLOSED DEFERRED | QA Contact: | GFS Bugs <gfs-bugs> | ||||
| Severity: | medium | Docs Contact: | |||||
| Priority: | medium | ||||||
| Version: | 4 | CC: | bstevens, rpeterso | ||||
| Target Milestone: | --- | ||||||
| Target Release: | --- | ||||||
| Hardware: | All | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2013-09-23 15:32:21 UTC | Type: | --- | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Attachments: |
|
||||||
Description of problem: ATIX uses nodeid as a reference for context dependent pathnames in their shared root configuration and setup. This patch adds the nodeid as an option to the context dependent pathname code. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: We had discussed this patch before and we'd come up with some issues. "o The nodeid in gfs is obtained from cman kernel module (and called "nodeX" where X is the nodeid) and thus makes gfs dependent on cman. This is not a good thing for non-clustered gfs deployment. Also, the node-id attribute is cluster-specific and doesn't really make sense in a non-clustered gfs installation. o Moreover, even in a clustered environment, the nodeids change over cluster-restarts. So, if a node was looking at one location for its files using @nodeid, after a cluster restart @nodeid may no longer point to that location." As I figured out later, the second point is a non-issue because we can force the nodes to have specific nodeids by explicitly declaring them in the cluster.conf. Do we still need this feature? Nodeids can be hard-wired to nodes in CCS. In fact we will insist on this for RHEL5 (ccs_tool has an subcommand to add them). done. added patch to RHEL4 and STABLE branches. The Red Hat Cluster Suite product is past end-of-life; closing. |
Created attachment 127991 [details] Adds nodeid symlink support to gfs