// ultimately we should be using a form of fetch obj that doesn't write to rados maybe?
string obj_path = bucket_info.bucket.name + "/" + key.name;
- ldout(store->ctx(),0) << "abhi: path=" << obj_path << dendl;
+ ldout(store->ctx(),0) << "AWS: path=" << obj_path << dendl;
// Don't try this at home, very hacky, probably need a proper aws client
// written so RESTResourceCR expects a JSON as a result, we workaround
if (retcode < 0) {
return set_cr_error(retcode);
}
- ldout(sync_env->cct,0) << "abhi: download complete, creating buckets"<< dendl;
bucket_name=aws_bucket_name(bucket_info);
if (bucket_created.find(bucket_name) == bucket_created.end()){
// // TODO: maybe do a head request for subsequent tries & make it configurable
yield {
//string bucket_name = aws_bucket_name(bucket_info);
+ ldout(sync_env->cct,0) << "AWS: creating bucket" << bucket_name << dendl;
bufferlist bl;
call(new RGWPutRawRESTResourceCR <int> (sync_env->cct, conf.conn,
sync_env->http_manager,
yield {
string path=aws_object_name(bucket_info, key);
- ldout(sync_env->cct,0) << "abhi sending request to " << conf.conn << " path" << path << dendl;
+ ldout(sync_env->cct,0) << "AWS creating object at path" << path << dendl;
call(new RGWPutRawRESTResourceCR<int> (sync_env->cct, conf.conn,
sync_env->http_manager,
path, nullptr,