Bug 13705 - master] bitbake and hashserve.sock left behind when ^C a build
Summary: master] bitbake and hashserve.sock left behind when ^C a build
Status: RESOLVED FIXED
Alias: None
Product: OE-Core
Classification: Build System, Metadata & Runtime
Component: core (show other bugs)
Version: 3.1
Hardware: x86 Multiple
: Medium+ normal
Target Milestone: 4.1 M1
Assignee: Armin Kuster
QA Contact:
URL:
Whiteboard:
Depends on:
Blocks:
 
Reported: 2019-12-13 18:14 UTC by Armin Kuster
Modified: 2022-04-28 16:53 UTC (History)
6 users (show)

See Also:
OS type for building Yocto: ---
Type of Regression: ---
Verified:
Documentation change: Don't know


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Armin Kuster 2019-12-13 18:14:45 UTC
I have these variables defined:

SOURCE_MIRROR_URL ?= "http://downloads.yoctoproject.org/mirror/sources/"
SSTATE_MIRRORS ?= "file://.* http://sstate.yoctoproject.org/PATH"


BB_SIGNATURE_HANDLER = "OEEquivHash"


The mirrors are down so there is a very long ETA. 

I control-c and bitbake.sock and hashserve.sock are left behind.

Next time bitbake is invoked, it times out and the cooker log indicates hashserve.sock exists.
Comment 1 Armin Kuster 2019-12-13 18:15:15 UTC
or I get this traceback

NOTE: Reconnecting to bitbake server...
NOTE: Retrying server connection (#8)... (Traceback (most recent call last):
  File "/home/akuster/oss/maint/poky/bitbake/lib/bb/main.py", line 455, in setup_bitbake
    server_connection = bb.server.process.connectProcessServer(sockname, featureset)
  File "/home/akuster/oss/maint/poky/bitbake/lib/bb/server/process.py", line 490, in connectProcessServer
    sock.connect(os.path.basename(sockname))
BlockingIOError: [Errno 11] Resource temporarily unavailable
)
ERROR: Unable to connect to bitbake server, or start one (server startup failures would be in bitbake-cookerdaemon.log)
Comment 2 Armin Kuster 2019-12-13 18:17:38 UTC
I think there are 3 problems here:

1) No good timeout for SSTATE_MIRROR
2) Traceback
3) Cleanup after cntl-c

I can open additional bugs for the other two.
Comment 3 Richard Purdie 2019-12-17 00:00:02 UTC
I suspect the files "left behind" still belong to a currently running bitbake server which isn't responsive.

Was there bitbake processes left behind from the first Ctrl+C'd build?
Comment 4 Armin Kuster 2019-12-19 15:44:53 UTC
we will create a new bug for the SSATE_MIRROR time and leave this as is.
Comment 5 Armin Kuster 2019-12-19 16:27:02 UTC
opened

https://bugzilla.yoctoproject.org/show_bug.cgi?id=13716 for the SSTATE_MIRROR issue.

This then becomes the cntl-c issue
Comment 6 Randy MacLeod 2022-04-11 22:24:05 UTC
Armin, We've fixed a few Control-C issues, could you check if this is still reproducible or if you prefer, one of us could just move it to 3.6-M1.
Comment 7 Randy MacLeod 2022-04-21 15:14:26 UTC
Is this still a problem?
Comment 8 Randy MacLeod 2022-04-28 14:56:46 UTC
Ping.
Comment 9 akuster 2022-04-28 16:53:32 UTC
works for me.