Re: (ITS#4050) syncrepl to back-hdb doesn't work

[quanah@stanford.edu]

--On Thursday, September 29, 2005 4:20 AM -0700 Howard Chu <hyc@symas.com> 
wrote:

> Do you remember the date/time just before the MODs started? We could
> start the slave with an explicit cookieCSN to make it rerun the changes
> and watch it under gdb. Really I guess we could start it with any
> relatively early CSN, it might re-add a few entries or whatnot but no big
> deal as long as we see some action.

I just recently reloaded the server, so I have the contextCSN from the time 
it was reloaded.  I modified the contextCSN value on the back-hdb slave to 
what it was at load time, and then restarted it under syncrepl.  It spent 
about 20 minutes being very busy back and forth with the master, with a net 
result of *no changes* happening on it.  It puts out exactly the same data 
it had before, which means the sync's are failing.

It also means we can debug this fairly easily once you are landed. ;)

--Quanah

--
Quanah Gibson-Mount
Principal Software Developer
ITSS/Shared Services
Stanford University
GnuPG Public Key: http://www.stanford.edu/~quanah/pgp.html

"These censorship operations against schools and libraries are stronger
than ever in the present religio-political climate. They often focus on
fantasy and sf books, which foster that deadly enemy to bigotry and blind
faith, the imagination." -- Ursula K. Le Guin