-
Notifications
You must be signed in to change notification settings - Fork 1.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
bcmsh is blocking bcmcmd, and no timeout either #1762
Comments
what's your expectation? |
we can only support one client at one time. The drivshell is not designed to support multiple shells. |
If only one shell is supported, if we start bcmsh, we should print the "warning" messages so people will be careful of using it as it could block regular CLI. Second, if someone runs bcmsh today and close the terminal, no timeout was done, bcmcmd etc will be blocked forever. We should enable time out to close bcmsh automatically, or at least bcmcmd related CLI should detect bcmsh and print messages to let user act. |
address in #1921 . add inactivity timer. bcmcmd has default timemout, it will not block forever. |
Based on my testing, there was no default timeout, I connect socat to the UNIX socket and kill the terminal, it never (at least hours) exits and no bcmcmd command can run after that unless you kill that socat session. |
can you elaborate, did you test #1921. |
The comments were for bcmsh testing before the changes in #1921. After the changes, it should be fine. You can close this. |
Signed-off-by: Vivek Reddy Karri <vkarri@nvidia.com> ce11545 [config reload] Removed job-mode for sonic.target restart (sonic-net#1820) f76f672 [fdbshow]: Fix typo in comment (sonic-net#1809) 17208a0 [ci]: Support PR coverage (sonic-net#1806) c2c2354 fix wrong code indent in sfputil (sonic-net#1808) 47a9a0f [portconfig] Validate duplicate speed value and interface type value (sonic-net#1745) f1086ee [sonic_installer]Add --skip-platform-check option for sonic_installer when image mismatch (sonic-net#1791) c007d65 [warm-reboot] Add new preboot health check: verify database integrity (sonic-net#1785) 41e31e8 Fix PatchApplier init order (sonic-net#1762)
0d538d3 [ci]: Support code diff coverage (#1834) 48887d1 [config] support for configuring muxcable to standby mode of operation (#1837) 2088a9a Provide support to install platform extensions (#1578) c97fe54 Add check_db_integrity script to setup.py (#1828) c0b9917 [debug dump util] COPP Module Added (#1670) 826311c [techsupport] Removed interactive option for docker commands and Improved Error Reporting (#1723) ce11545 [config reload] Removed job-mode for sonic.target restart (#1820) f76f672 [fdbshow]: Fix typo in comment (#1809) 17208a0 [ci]: Support PR coverage (#1806) c2c2354 fix wrong code indent in sfputil (#1808) 47a9a0f [portconfig] Validate duplicate speed value and interface type value (#1745) f1086ee [sonic_installer]Add --skip-platform-check option for sonic_installer when image mismatch (#1791) c007d65 [warm-reboot] Add new preboot health check: verify database integrity (#1785) 41e31e8 Fix PatchApplier init order (#1762) 2416175 [config reload] Fix config reload failure due to sonic.target job cancellation (#1814) 2b12aad [portstat, intfstat] added rates and utilization (#1750) 26e700a [debug dump util] Techsupport addition (#1669) 9f2326e [debug dump util] Base Skeleton and Click Class added (#1668)
…rt auto negotiation (sonic-net#1762) - What I did Update the speed used for headroom calculating according to port auto-negotiation change. The speed taken into the calculation is now called effective speed in the code. - Why I did it One logic was to take the configured speed into account when calculating the headroom because it's the maximum possible speed at which a port runs. As port auto-negotiation supported, this logic needs to be updated accordingly: The maximum possible speed is no longer the configured one with auto-negotiation enabled. So the speed taken into account should be: If port auto-negotiation is enabled: maximum speed in adv_speeds if it is configured maximum speed in supported_speeds otherwise Otherwise: the configured speed - How I verified it Run regression test and vs test. Signed-off-by: Stephen Sun stephens@nvidia.com
#### What I did Otherwise PatchSorter will see both config_wrapper, patch_wrapper as None. #### How I did it #### How to verify it #### Previous command output (if the output of a command-line utility has changed) #### New command output (if the output of a command-line utility has changed)
Description
/var/run/sswsyncd/sswsyncd.socket seems to only allow one client, this is problematic especially when we have bcmsh enabled. If some one is using bcmsh, and bcmcmd will hang there. Ever worse, bcmsh does not time-out, so bcmcmd could be blocked there forever until you kill the bcmsh.
Steps to reproduce the issue:
run bcmsh
a# bcmsh
Press Enter to show prompt.
Press Ctrl+C to exit.
drivshell>
bcmcmd ps
<== stuck here
Describe the results you received:
Describe the results you expected:
Additional information you deem important (e.g. issue happens only occasionally):
The text was updated successfully, but these errors were encountered: