On Fri, May 18, 2018 at 6:34 AM, Xinying Song <songxinying.ftd@xxxxxxxxx> wrote: > Hi, guys > > RGWDataSyncShardCR will add a lock on its status oid before doing sync. The > lock will re-lock itself every half duration. With a positive duration it > will be released forcibly if time expires when another lock request from > different client reached. Since the re-lock process is executed in the same > thread with coroutine, the re-lock maybe lagging. In this situation, both > RGWDataSyncShardCRs will think it owns the lock and continues to do sync. > > So my question is why not use a zero duration and let RGWDataSyncShardCR > give up a lock on its own (it will indeed release the lock)? > Because if rgw goes down uncleanly, there will be nothing to break that lock, and sync won't really work anymore. Yehuda > Thanks, > Xinying. > -- > To unsubscribe from this list: send the line "unsubscribe ceph-devel" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
