amazon s3 - Cloud Storage Appending and Logging -

what cloud service combination allow storage of logs n parts such cheap , easy to:

1. append new entries log
2. request parts i -> j log, or @ least parts i -> n
3. have requests above consistent (i.e. reading log after user appends entry n + 1 results in entry n + 1 being available)
4. log entries potentially quite large (10's 100's of mb) imagine user uploads picture service log entry i'm storing [log , user data | picture content]. while can happily store log , user data separately picture content, when retrieve log need efficiently able both sections.

some research , thoughts have:

• s3/google cloud storage not seem allow modifying object data, although google allows compose function might useful.
• google cloud storage , windows azure storage consistent, s3 not although might able use db deal consistency issues here.
• microsoft azure might best bet, since blobs seem allow manipulating individual pages/blocks. however, i'm not sure whether using blobs satisfy above requirements if i'm ok caps on maximum blob size. example, satisfying requirement 2 seems difficult without having separate location (blob/db) correlates log number blob number.

answers suggesting have come far being close optimal fine, wondering whether have missed strategy regarding these storage providers or provider i've entirely missed. additionally, if i'm stuck using db data storage anyway it'd interesting if minor tweaking allow work efficiently across multiple cloud providers cost , redundancy purposes, although may pipe dream.

you can consider storing every log entry separate blob in azure storage. solution;

1. creates new blob
2. reads blobs i->j
3. entire block blob content becomes readable committed single api call (put block list)
4. a single block blob can store 50k*4mb, can store entire log entry in 1 blob.

moreover, each blob has own partition in azure storage. blobs can therefore distributed across many servers in order scale out access them.