google compute engine - Persistent disk reliability and backups in GCE -
(my apologies being off-so question, seems gce questions tend loess tightly connected programming.)
i designing data acquisition program run on gce. data collected onto persistent disk. data cannot afford lose, need know reliability of persistent disks. have been able find 3 pieces of information:
persistent disks have built-in redundancy protect data against equipment failure , remain available through datacenter maintenance events. instances, free of local storage, can moved google live migration newer hardware without intervention. allows google datacenters maintained @ highest level; software, hardware, , facilities can continually updated ensure excellent performance , reliability cloud-based services. [google]
google compute engine uses redundant, industry-standard mechanisms protect persistent disk users data corruption , sophisticated attacks against data integrity. [google]
google persistent disk never return erroneous data, instead there io error. [i cannot find 1 right now, remember reading of google's docs, take pinch salt]
the cloud storage comes reliability numbers, there information persistent disks? without knowing reliability estimates difficult choose backup regime. opposite ends of continuum using hot backups real time synchronization , taking regular deltas of data (the data append-only). in latter case recovery takes longer , involve manual bit-stitching. can afforded, if mtbf high enough.
i not worried brief downtimes, worried data corruption. system running on linux+ext4, should resilient against unplanned downtime.
you can go through following blog post gives more info persistence disk https://cloud.google.com/developers/articles/compute-engine-disks-price-performance-and-persistence
Comments
Post a Comment