Closed Bug 988199 Opened 11 years ago Closed 7 years ago

[MTBF] get_about_memory.py causing running out of space in Buri

Categories

(Firefox OS Graveyard :: MTBF, defect)

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(Not tracked)

RESOLVED WONTFIX

People

(Reporter: wachen, Unassigned)

References

Details

After getting memory report about 20 times (once/hour or so), the temp folder is getting too large that the device stopped for no space(ran out of space). 

PS. It was a bug found like last year when nuwa was not in. It might not be able to reproduce due to another blocking bug

I will try to get the Jenkins report link posted here.
Blocks: MTBF-meta
Kyle, sorry for the bothering. Per our discussion, I am opening this bug keep the bug in record. I will update this soon.
Depends on: 983563
http://mtbf-1.corp.tpe1.mozilla.com:8080/job/mtbf.logcat.multi/label=master/7/console

E/GeckoConsole(  136): [JavaScript Error: "NS_ERROR_FILE_NO_DEVICE_SPACE: Component returned failure code: 0x80520010 (NS_ERROR_FILE_NO_DEVICE_SPACE) [nsIFileOutputStream.write]" {file: "chrome://marionette/content/marionette-listener.js" line: 2048}]

I believe this is what stopped the phone
Blocks: MTBF-Memory
No longer blocks: MTBF-meta
Summary: [Memory Report] get_about_memory.py causing running out of space in Buri → [MTBF] get_about_memory.py causing running out of space in Buri
The profiler originally had this problem. It was modified to remove previous profiles before creating new ones.
Interesting. I will pull the most up-to-date get_about_memory things to test.
Component: Performance → MTBF
Walter can you still reproduce this? More recent versions of |get_about_memory.py| should clear out the logs after retrieval. 

It's possible that in error scenarios that we're not properly clearing out dumped logs (such as a process dying while trying to retrieve a memory report).
Flags: needinfo?(wachen)
We stopped using it for quite a long time though. So, no reproduction made.

What kind of dump logs do you need? How do I fetch them? Is it like a crash report or...?
Flags: needinfo?(wachen)
(In reply to Walter Chen[:ypwalter][:wachen] from comment #6)
> We stopped using it for quite a long time though. So, no reproduction made.
> 
> What kind of dump logs do you need? How do I fetch them? Is it like a crash
> report or...?

We just need to retest and see if the device runs out of tmp space. If that happens any python exceptions from the script would be useful.
Firefox OS is not being worked on
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.