rpc.gssd memory leak
Bug #988055 reported by
Steve Atwell
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| nfs-utils (Ubuntu) |
Fix Released
|
High
|
Adam Stokes | ||
| Lucid |
Won't Fix
|
High
|
Dimitri John Ledkov | ||
Bug Description
rpc.gssd appears to be leaking a small amount of memory every time it's used to authenticate a kerberized nFS mount. I can reproduce by running:
while sleep 1; do mount -o sec=krb5 nfsserver:
While running this in a loop, you can observe from rpc.gssd's /proc/<pid>/status file that its VmRSS grows indefinitely, generally by about 4 kB per mount request. On some of our servers that have a high turn-over rate of kerberized NFS mounts, we've seen the rpc.gssd process grow to more than 600 MB!
This happens on lucid with nfs-common 1:1.2.0-4ubuntu4.2.
I haven't been able to reproduce on precise with nfs-common 1:1.2.5-3ubuntu3, but this is still an important bug for lucid.
| Changed in nfs-utils (Ubuntu): | |
| status: | New → Incomplete |
| status: | Incomplete → Opinion |
| status: | Opinion → Invalid |
| status: | Invalid → Won't Fix |
| status: | Won't Fix → Confirmed |
Revision history for this message
| Adam Stokes (adam-stokes) wrote | #1 |
| Changed in nfs-utils (Ubuntu): | |
| importance: | Undecided → High |
| assignee: | nobody → Adam Stokes (adam-stokes) |
| Changed in nfs-utils (Ubuntu): | |
| status: | Confirmed → Fix Released |
| Changed in nfs-utils (Ubuntu Lucid): | |
| status: | New → Triaged |
| importance: | Undecided → High |
| assignee: | nobody → Adam Stokes (adam-stokes) |
| Changed in nfs-utils (Ubuntu Lucid): | |
| assignee: | Adam Stokes (adam-stokes) → Dmitrijs Ledkovs (xnox) |
Revision history for this message
| Rolf Leggewie (r0lf) wrote | #2 |
| Changed in nfs-utils (Ubuntu Lucid): | |
| status: | Triaged → Won't Fix |
To post a comment you must log in.
Hi,
I am still looking into the root cause of the problem. Thank you for your patience.
Adam