"hg push" hangs for one minute (exactly one minute)

Ernie Rael err at raelity.com
Thu Apr 28 13:24:48 UTC 2022


On 4/28/22 6:16 AM, Scott Palmer wrote:
>
> Scott
>
>> On Apr 28, 2022, at 9:01 AM, Ernie Rael <err at raelity.com> wrote:
>>
>> On 4/28/22 12:59 AM, Anton Shestakov wrote:
>>> There are also --debug and --profile, maybe they can help. Although
>>> some parts of the push process may not have debug logging, and
>>> profiling is for when you wish to delve into source code.
>>>
>>> I suspect there's something in the setup of the network or network
>>> file system, because 1 minute looks like a net IO timeout. But it
>>> could also be a timeout for waiting for a file lock, --profile should
>>> show it if that's the case.
>>>
>> Thanks for the options. Bingo.
>>
>>
>> Twice it does:
>>      ... pause for 30 seconds ...
>>      evoext-cache: error while saving new data: database is locked
> So if you disable the evolve extension does the pause go away?
> I wonder if this is evolve-related why it hits the issue while the rest of Mercurial doesn’t.

Another bingo.

pushing to /z/repos/jvi-dev/jvi
obsolete feature not enabled but 804 markers found!
searching for changes
no changes found

real    0m0.201s
user    0m0.147s
sys    0m0.008s


> Scott
>
>




More information about the Mercurial mailing list