which OS are you using? it does not look at the timestamps to decide if the index is in sync . It looks at the index version only.
BTW can you just hit the master withe url and paste the response here http://<masterhost>:<port>/solr/replication?command=filelist On Thu, Apr 23, 2009 at 11:53 AM, Jian Han Guo <jian...@gmail.com> wrote: > That's right. The timestamp of files on the slave side are all Dec 31 1969, > so it looks the timestamp was not set (and therefore it is zero). The ones > on the master side are all correct. Nevertheless, solr seems being able to > recognize that master and slave are in sync after replication. Don't know > how it does that. > > I haven't check if the two machines are in sync, but even if they are not, > the timestamp should not be Dec 31, 1969, I think. > > Thanks, > > Jianhan > > > > 2009/4/22 Noble Paul നോബിള് नोब्ळ् <noble.p...@gmail.com> > >> Let me assume that you are using the in-inbuilt replication. >> >> The replication ties to set the timestamp of all the files same as >> that of the files in the master. just cross check. >> >> On Thu, Apr 23, 2009 at 6:57 AM, Jian Han Guo <jian...@gmail.com> wrote: >> > Hi, >> > >> > I am using nightly build on 4/22/2009. Replication works fine, but the >> files >> > inside index directory on slave side all have old timestamp: Dec 31 >> 1969. >> > Is this a known issue? >> > >> > Thanks, >> > >> > Jianhan >> > >> >> >> >> -- >> --Noble Paul >> > -- --Noble Paul
