[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v4 01/15] xenstore: add support for a retry open limit on xenstored

On Mon, May 12, 2014 at 8:37 AM, Ian Jackson <Ian.Jackson@xxxxxxxxxxxxx> wrote:
> Jacek Konieczny writes ("Re: [PATCH v4 01/15] xenstore: add support for a 
> retry open limit on xenstored"):
>> On 05/12/14 16:18, Ian Campbell wrote:
>> > On Mon, 2014-05-12 at 14:50 +0100, Ian Jackson wrote:
>> >> I'm afraid I fundamentally disagree with this timeout/polling based
>> >> approach.
>> >
>> > As for the retry -- I suppose xenstored forks before it is really ready?
>> I guess this should be fixed in xenstored instead of hacking
>> work-arounds in the client. When daemon start succeeds then the socket
>> should be available and working. If it failed, no retry will help. If
>> the daemon is still starting although the start command exited â than
>> this sounds like a bug.
> Right.
> I think moving this retry from a shell script into the client code is
> going in the wrong direction.

At this point the retry option can not only be used for opening the
socket but for anything so its a bit different now. In terms of design
of issues specifically to opening the socket -- I believe using
systemd is actually a good step forward to move away from all this
legacy bullshit.

Either way, I'm happy to drop this patch from my series, originally I
just wanted to avoid nasty shell code and wanted to try to share code.
You guys let me know if I should drop it, for now I'll keep it in the


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.