Hi Otis, I think these are related problems but giving the ability to enforce a majority quorum among the total replica set for a shard is not the same as hinted handoff in the Cassandra sense. Cass's hinted handed allows you to say it's ok to send the write somewhere and somehow it'll make its way to the correct node, eventually. But, from the docs "A hinted write does not count towards ConsistencyLevel requirements of ONE, QUORUM, or ALL." It's useful for environments that need extreme write-availability.
Today, SolrCloud uses a consistency level of ALL, where ALL is based on the number of *active* replicas for a shard, which could be as small as one. What I'm proposing is to give a knob to allow a SolrCloud user to enforce a consistency level of QUORUM, where QUORUM is based on the entire replica set (down + active replicas) and not just active replicas as it is today. However, we'll need a better vocabulary for this because in my scenario, QUORUM is stronger than ALL which will confuse even the most seasoned of distributed systems engineers ;-) Cheers, Tim On Tue, Nov 19, 2013 at 9:25 PM, Otis Gospodnetic <otis.gospodne...@gmail.com> wrote: > Btw. isn't the situation Timothy is describing what hinted handoff is all > about? > > http://wiki.apache.org/cassandra/HintedHandoff > http://www.datastax.com/dev/blog/modern-hinted-handoff > > Check this: > http://www.jroller.com/otis/entry/common_distributed_computing_routines > > Otis > -- > Performance Monitoring * Log Analytics * Search Analytics > Solr & Elasticsearch Support * http://sematext.com/ > > > On Tue, Nov 19, 2013 at 1:58 PM, Mark Miller <markrmil...@gmail.com> wrote: > >> Mostly a lot of other systems already offer these types of things, so they >> were hard not to think about while building :) Just hard to get back to a >> lot of those things, even though a lot of them are fairly low hanging >> fruit. Hardening takes the priority :( >> >> - Mark >> >> On Nov 19, 2013, at 12:42 PM, Timothy Potter <thelabd...@gmail.com> wrote: >> >> > You're thinking is always one-step ahead of me! I'll file the JIRA >> > >> > Thanks. >> > Tim >> > >> > >> > On Tue, Nov 19, 2013 at 10:38 AM, Mark Miller <markrmil...@gmail.com> >> wrote: >> > >> >> Yeah, this is kind of like one of many little features that we have just >> >> not gotten to yet. I’ve always planned for a param that let’s you say >> how >> >> many replicas an update must be verified on before responding success. >> >> Seems to make sense to fail that type of request early if you notice >> there >> >> are not enough replicas up to satisfy the param to begin with. >> >> >> >> I don’t think there is a JIRA issue yet, fire away if you want. >> >> >> >> - Mark >> >> >> >> On Nov 19, 2013, at 12:14 PM, Timothy Potter <thelabd...@gmail.com> >> wrote: >> >> >> >>> I've been thinking about how SolrCloud deals with write-availability >> >> using >> >>> in-sync replica sets, in which writes will continue to be accepted so >> >> long >> >>> as there is at least one healthy node per shard. >> >>> >> >>> For a little background (and to verify my understanding of the process >> is >> >>> correct), SolrCloud only considers active/healthy replicas when >> >>> acknowledging a write. Specifically, when a shard leader accepts an >> >> update >> >>> request, it forwards the request to all active/healthy replicas and >> only >> >>> considers the write successful if all active/healthy replicas ack the >> >>> write. Any down / gone replicas are not considered and will sync up >> with >> >>> the leader when they come back online using peer sync or snapshot >> >>> replication. For instance, if a shard has 3 nodes, A, B, C with A being >> >> the >> >>> current leader, then writes to the shard will continue to succeed even >> >> if B >> >>> & C are down. >> >>> >> >>> The issue is that if a shard leader continues to accept updates even if >> >> it >> >>> loses all of its replicas, then we have acknowledged updates on only 1 >> >>> node. If that node, call it A, then fails and one of the previous >> >> replicas, >> >>> call it B, comes back online before A does, then any writes that A >> >> accepted >> >>> while the other replicas were offline are at risk to being lost. >> >>> >> >>> SolrCloud does provide a safe-guard mechanism for this problem with the >> >>> leaderVoteWait setting, which puts any replicas that come back online >> >>> before node A into a temporary wait state. If A comes back online >> within >> >>> the wait period, then all is well as it will become the leader again >> and >> >> no >> >>> writes will be lost. As a side note, sys admins definitely need to be >> >> made >> >>> more aware of this situation as when I first encountered it in my >> >> cluster, >> >>> I had no idea what it meant. >> >>> >> >>> My question is whether we want to consider an approach where SolrCloud >> >> will >> >>> not accept writes unless there is a majority of replicas available to >> >>> accept the write? For my example, under this approach, we wouldn't >> accept >> >>> writes if both B&C failed, but would if only C did, leaving A & B >> online. >> >>> Admittedly, this lowers the write-availability of the system, so may be >> >>> something that should be tunable? Just wanted to put this out there as >> >>> something I've been thinking about lately ... >> >>> >> >>> Cheers, >> >>> Tim >> >> >> >> >> >>
