Also the WalObserver hooks were also used to ensure that the append to WAL happens thro them.
Regards Ram On Wed, Dec 5, 2012 at 4:58 PM, ramkrishna vasudevan < [email protected]> wrote: > Thanks Anoop for the reply. I was planning to reply if not from you. > > Regards > Raml > > > On Wed, Dec 5, 2012 at 4:34 PM, Anoop Sam John <[email protected]> wrote: > >> Hi Jan >> Yes we guarentee the consistency between user table and index >> table. The put operation will be handled as a transactional way so as to >> make sure the data is added to both tables or reverted back from both. Some >> new CP hooks we have added for this obviously. >> >> -Anoop- >> ________________________________________ >> From: Jan Van Besien [[email protected]] >> Sent: Wednesday, December 05, 2012 12:54 AM >> To: [email protected] >> Subject: Re: HBase - Secondary Index >> >> Hi, >> >> On 12/04/2012 09:10 AM, Anoop Sam John wrote: >> > I would like to hear what others think on this. :) >> >> I found it interesting to read your approach on how the indexes can be >> used to speed up existing scan operations. >> >> I couldn't find anything in your presentation though about whether your >> implementation makes any guarantees to ensure the source table and the >> index table are always (eventually) in sync. >> >> What if data is inserted in the source table and then the region server >> crashes (before the coprocessor is executed)? Will the index be out of >> sync? Do you have a mechanisme in place to detect and restore this >> situation? >> >> thanks >> Jan >> > >
