SO bug causing crash

Getting back to business, I don't know which bug you're referring to.
But there was a shared object bug fixed in the few builds between
1.5.2r138 and 141.  Not sure if it addresses the problem you are facing.

But if memory serves me, and I'll be the first to admit that it doesn't
serve me well, but I believe this bug could result in a crash.  But I
don't believe it is specific to proxied shared objects.

I'm not sure if this addresses your problem; I hope it does.  Also,
while I don't doubt there are bugs in proxied shared objects, I do find
it hard to believe they are not usable.  I know Justin Watkins has a
huge distributed setup built using 1.5.2 and proxied streams and SO's.

I'm sure that's not the answer you want.  But somebody else had said
that Asa was the point of contact on this issue.  So I'll let him look
into it when he gets back from vacation since he may have some context
on it. =20

But one last point, and I promise this will be the last thing I have to
say on the matter.  This list is a great source of information for you,
the FCS/FMS developers, as well as for us.  We gladly monitor this list
to help you guys out when possible.  But for any real issues, if you
want an official response, you should be going thru the proper
Macromedia support channels.  If that is not working, you should take it
up with Macromedia support.  We do our best to respond to questions and
problems that are reported on this list.  But make no mistake, it is not
in our job description to do so, and we are certainly not required to be
a slave to this list and to have every issue and problem reported on
this list burned into memory.  Each one of us monitors this list often
on our own time and you often get speedy responses from us, even in the
wee hours of the night.  I've never mouthed off to any one of you
before, and as Pritham said, we would expect the same mutual respect.
We like to say that no question is a dumb one, but we've also had our
fair share of not so intelligent questions, which we gladly respond to
in a non-condescending and respectful manner.  If you can't do the same,
don't expect us to bend over backwards if you need help.  So I'd like to
apologize to Jayson for calling him a prick.  It was said in a moment of
anger.  So let's get back to business as usual and let the flaming end
here.

Regards,
Ed

> -----Original Message-----
> From: flashcomm-bounces-1Ss2GqJETD3yZ38Mhd3e/9ZfFG6BLHNm@public.gmane.org=
20
> [mailto:flashcomm-bounces-1Ss2GqJETD3yZ38Mhd3e/9ZfFG6BLHNm@public.gman
e.org] On Behalf Of JesterXL
> Sent: Thursday, November 17, 2005 10:43 PM
> To: FlashComm Mailing List
> Subject: Re: [FlashComm] updater 3 question
>=20 
> I'm showing
> the bandwidth maxed out at 40mb and the connections at ~8400.=20
> Right now, this is all running on a single (rather=20
> impressive) server, becuse load balancing on 1.5 is=20
> impossible. I say it's impossible because I happen to know=20
> that there is a known bug in remote shared objects that=20
> prevents you from actually implementing a usable load=20
> balancing solution. The comm server service crashes under any=20
> real load, effectively negating the purpose of load=20
> balancing. Of course the bug was immediately addressed in=20
> FMS, but no patch was released for 1.5, to which my response=20
> was "That's fine, I'll just make them upgrade". Now that's no=20
> longer an option. 
>=20 
> Because I need load balancing, and since they apparently=20
> still don't have proxied SOs working properly, the only way=20
> to do it is with the Origin/Edge setup. 
>=20 
>  It was discussed on this list just a couple months ago. It's=20
> a bug in the constructor for proxied shared objects that=20
> crashes the comm server service. It's a problem if you want=20
> your comm server to have more than about
> 12 hours of uptime. Under a load, our servers would crash=20
> every few minutes.
>=20
>     The part that irritates me about this bug is that when it=20
> was identified, it was immediately fixed in FMS2, but FCS1.5=20
> was not patched, essentially saying to us "if you want load=20
> balancing, upgrade to FMS", which would be fine if it=20
> wouldn't cost us 25 times as much to do the same thing.
> Forcing us to upgrade at a 2500% cost increase to address a=20
> known bug in their software is just wrong. 
> ----- Original Message -----
> From: "Edward Chan" <echan-14osZcCZf762oZ/6fjIToQ@public.gmane.org>
> To: "FlashComm Mailing List" <flashcomm-1Ss2GqJETD3yZ38Mhd3e/9ZfFG6BLHNm@p
ublic.gmane.org>
> Sent: Friday, November 18, 2005 12:54 AM
> Subject: RE: [FlashComm] updater 3 question
>=20
>=20
> Sorry, but I still don't understand what the problem is?  How are you
> doing the load balancing?  And what exactly are you doing that is
> causing FCS to crash?  Do you mean you have multiple FCS=20
> servers behind
> a load balancer, and when the load balancer sends a=20
> connection to one of
> the FCS servers, it crashes?
>=20
> Ed
>=20 
> lazy to go 
> http://www.macromedia.com/support/f...loads_updaters. 
=3D-----------------------------------------------------------[vbcol=seagreen] 
=3D-----------------------------------------------------------[vbcol=seagreen] 
>=20
> Supported by Fig Leaf Software - http://www.figleaf.com
>=20
> To change your subscription options or search the archive:
> http://chattyfig.figleaf.com/mailma...fo/flashcomm=20
>=20
>=20
> =3D-----------------------------------------------------------
> Supported by Fig Leaf Software - http://www.figleaf.com
> =3D-----------------------------------------------------------
>=20
> To change your subscription options or search the archive:
> http://chattyfig.figleaf.com/mailman/listinfo/flashcomm
>=20

=-----------------------------------------------------------
Supported by Fig Leaf Software - http://www.figleaf.com
=-----------------------------------------------------------

To change your subscription options or search the archive:
http://chattyfig.figleaf.com/mailman/listinfo/flashcomm