Closed Bug 733078 Opened 12 years ago Closed 9 years ago

hg share called multiple times from hgtool.py

Categories

(Release Engineering :: General, defect, P4)

x86_64
Windows 7

Tracking

(Not tracked)

RESOLVED INCOMPLETE

People

(Reporter: joey, Unassigned)

Details

(Whiteboard: [hg])

Might be able to save some shell/command overhead by remembering detected state

=================================================================================
builder: try-win64
slave: w64-ix-slave34
starttime: 1330966046.17
results: exception (4)
buildid: 20120305084743
builduid: 118e3e49e7334ca9a7dacb559173c0d1
revision: d1fa189b44ce

http://ftp.mozilla.org/pub/mozilla.org/firefox/try-builds/jarmstrong@mozilla.com-d1fa189b44ce/try-win64/try-win64-build5706.txt.gz
=================================================================================

Checking if share extension works
command: hg help share
hg: unknown command 'share'

[snip]

Disabling sharing since share extension doesn't seem to work (3)
Attempting to clone from mirrors
command: hg clone -r d1fa189b44cea5f8110c2f4789bffaee66ab132b
abort: unknown revision 'd1fa189b44cea5f8110c2f4789bffaee66ab132b'!

[snip]

Back to 'hg share again'

Pulling from mirrors failed; falling back to http://hg.mozilla.org/try
Checking if share extension works
hg: unknown command 'share'


ps>  Is "checking out from mirrors" considered a valid op for a try submission ?  Perhaps caching a value of the patch submitted to try.  If try patches are cached on a mirror to preserve state see bug #733050
(In reply to Joey Armstrong [:joey] from comment #0)
> Might be able to save some shell/command overhead by remembering detected
> state
> 
> =============================================================================
> ====
> builder: try-win64
> slave: w64-ix-slave34
> starttime: 1330966046.17
> results: exception (4)
> buildid: 20120305084743
> builduid: 118e3e49e7334ca9a7dacb559173c0d1
> revision: d1fa189b44ce
> 
> http://ftp.mozilla.org/pub/mozilla.org/firefox/try-builds/jarmstrong@mozilla.
> com-d1fa189b44ce/try-win64/try-win64-build5706.txt.gz
> =============================================================================
> ====
> 
> Checking if share extension works
> command: hg help share
> hg: unknown command 'share'
> 
> [snip]
> 
> Disabling sharing since share extension doesn't seem to work (3)
> Attempting to clone from mirrors
> command: hg clone -r d1fa189b44cea5f8110c2f4789bffaee66ab132b
> abort: unknown revision 'd1fa189b44cea5f8110c2f4789bffaee66ab132b'!
> 
> [snip]
> 
> Back to 'hg share again'
> 
> Pulling from mirrors failed; falling back to http://hg.mozilla.org/try
> Checking if share extension works
> hg: unknown command 'share'
> 
> 
> ps>  Is "checking out from mirrors" considered a valid op for a try
> submission ?  Perhaps caching a value of the patch submitted to try.  If try
> patches are cached on a mirror to preserve state see bug #733050

Yes, pulling from mirrors is a valid thing to try for try. Can you tell if the multiple checks for 'hg share' are in the same invocation of 'hgtool' or not?
Severity: normal → minor
Priority: -- → P4
Summary: tools/buildfarm/utils/retry.py -vs- win64 → hg share called multiple times from hgtool.py
Whiteboard: [hg]
Product: mozilla.org → Release Engineering
Status: NEW → RESOLVED
Closed: 9 years ago
Resolution: --- → INCOMPLETE
Component: General Automation → General
You need to log in before you can comment on or make changes to this bug.