This PR also adds some comments and simplifies
the code. Primary handling is done to ensure
that we make sure to honor cached buffer.
Added unit tests as well
Fixes#7141
- New parser written from scratch, allows easier and complete parsing
of the full S3 Select SQL syntax. Parser definition is directly
provided by the AST defined for the SQL grammar.
- Bring support to parse and interpret SQL involving JSON path
expressions; evaluation of JSON path expressions will be
subsequently added.
- Bring automatic type inference and conversion for untyped
values (e.g. CSV data).
This PR supports iam and bucket policies to have
policy variable replacements in resource and
condition key values.
For example
- ${aws:username}
- ${aws:userid}
When auto-encryption is turned on, we pro-actively add SSEHeader
for all PUT, POST operations. This is unusual for V2 signature
calculation because V2 signature doesn't have a pre-defined set
of signed headers in the request like V4 signature. According to
V2 we should canonicalize all incoming supported HTTP headers.
Make sure to validate signatures before we mutate http headers
Before this change the CopyObjectHandler and the CopyObjectPartHandler
both looked for a `versionId` parameter on the `X-Amz-Copy-Source` URL
for the version of the object to be copied on the URL unescaped version
of the header. This meant that files that had question marks in were
truncated after the question mark so that files with `?` in their
names could not be server side copied.
After this change the URL unescaping is done during the parsing of the
`versionId` parameter which fixes the problem.
This change also introduces the same logic for the
`X-Amz-Copy-Source-Version-Id` header field which was previously
ignored, namely returning an error if it is present and not `null`
since minio does not currently support versions.
S3 Docs:
- https://docs.aws.amazon.com/AmazonS3/latest/API/RESTObjectCOPY.html
- https://docs.aws.amazon.com/AmazonS3/latest/API/mpUploadUploadPartCopy.html
When source is encrypted multipart object and the parts are not
evenly divisible by DARE package block size, target encrypted size
will not necessarily be the same as encrypted source object.
This PR adds pass-through, single encryption at gateway and double
encryption support (gateway encryption with pass through of SSE
headers to backend).
If KMS is set up (either with Vault as KMS or using
MINIO_SSE_MASTER_KEY),gateway will automatically perform
single encryption. If MINIO_GATEWAY_SSE is set up in addition to
Vault KMS, double encryption is performed.When neither KMS nor
MINIO_GATEWAY_SSE is set, do a pass through to backend.
When double encryption is specified, MINIO_GATEWAY_SSE can be set to
"C" for SSE-C encryption at gateway and backend, "S3" for SSE-S3
encryption at gateway/backend or both to support more than one option.
Fixes#6323, #6696
minio-java tests were failing under multiple places when
auto encryption was turned on, handle all the cases properly
This PR fixes
- CopyObject should decrypt ETag before it does if-match
- CopyObject should not try to preserve metadata of source
when rotating keys, unless explicitly asked by the user.
- We should not try to decrypt Compressed object etag, the
potential case was if user sets encryption headers along
with compression enabled.
This commit adds an auto-encryption feature which allows
the Minio operator to ensure that uploaded objects are
always encrypted.
This change adds the `autoEncryption` configuration option
as part of the KMS conifguration and the ENV. variable
`MINIO_SSE_AUTO_ENCRYPTION:{on,off}`.
It also updates the KMS documentation according to the
changes.
Fixes#6502
This PR implements one of the pending items in issue #6286
in S3 API a user can request CSV output for a JSON document
and a JSON output for a CSV document. This PR refactors
the code a little bit to bring this feature.
This refactor brings a change which allows
targets to be added in a cleaner way and also
audit is now moved out.
This PR also simplifies logger dependency for auditing
To conform with AWS S3 Spec on ETag for SSE-S3 encrypted objects,
encrypt client sent MD5Sum and store it on backend as ETag.Extend
this behavior to SSE-C encrypted objects.
This improves the performance of certain queries dramatically,
such as 'count(*)' etc.
Without this PR
```
~ time mc select --query "select count(*) from S3Object" myminio/sjm-airlines/star2000.csv.gz
2173762
real 0m42.464s
user 0m0.071s
sys 0m0.010s
```
With this PR
```
~ time mc select --query "select count(*) from S3Object" myminio/sjm-airlines/star2000.csv.gz
2173762
real 0m17.603s
user 0m0.093s
sys 0m0.008s
```
Almost a 250% improvement in performance. This PR avoids a lot of type
conversions and instead relies on raw sequences of data and interprets
them lazily.
```
benchcmp old new
benchmark old ns/op new ns/op delta
BenchmarkSQLAggregate_100K-4 551213 259782 -52.87%
BenchmarkSQLAggregate_1M-4 6981901985 2432413729 -65.16%
BenchmarkSQLAggregate_2M-4 13511978488 4536903552 -66.42%
BenchmarkSQLAggregate_10M-4 68427084908 23266283336 -66.00%
benchmark old allocs new allocs delta
BenchmarkSQLAggregate_100K-4 2366 485 -79.50%
BenchmarkSQLAggregate_1M-4 47455492 21462860 -54.77%
BenchmarkSQLAggregate_2M-4 95163637 43110771 -54.70%
BenchmarkSQLAggregate_10M-4 476959550 216906510 -54.52%
benchmark old bytes new bytes delta
BenchmarkSQLAggregate_100K-4 1233079 1086024 -11.93%
BenchmarkSQLAggregate_1M-4 2607984120 557038536 -78.64%
BenchmarkSQLAggregate_2M-4 5254103616 1128149168 -78.53%
BenchmarkSQLAggregate_10M-4 26443524872 5722715992 -78.36%
```
In many situations, while testing we encounter
ErrInternalError, to reduce logging we have
removed logging from quite a few places which
is acceptable but when ErrInternalError occurs
we should have a facility to log the corresponding
error, this helps to debug Minio server.
This commit adds key-rotation for SSE-S3 objects.
To execute a key-rotation a SSE-S3 client must
- specify the `X-Amz-Server-Side-Encryption: AES256` header
for the destination
- The source == destination for the COPY operation.
Fixes#6754
This PR adds support
- Request query params
- Request headers
- Response headers
AuditLogEntry is exported and versioned as well
starting with this PR.
Execute method in s3Select package makes a response.WriteHeader call.
Not calling it again in SelectObjectContentHandler function in case of
error in s3Select.Execute call.
On a heavily loaded server, getBucketInfo() becomes slow,
one can easily observe deleting an object causes many
additional network calls.
This PR is to let the underlying call return the actual
error and write it back to the client.
Current master didn't support CopyObjectPart when source
was encrypted, this PR fixes this by allowing range
CopySource decryption at different sequence numbers.
Fixes#6698
This PR fixes
- The target object should be compressed even if the
source object is not compressed.
- The actual size for an encrypted object should be the
`decryptedSize`
This commit fixes a wrong assignment to `actualPartSize`.
The `actualPartSize` for an encrypted src object is not `srcInfo.Size`
because that's the encrypted object size which is larger than the
actual object size. So the actual part size for an encrypted
object is the decrypted size of `srcInfo.Size`.
Without this fix we have room for two different type of
errors.
- Source is encrypted and we didn't provide any source encryption keys
This results in Incomplete body error to be returned back to the client
since source is encrypted and we gave the reader as is to the object
layer which was of a decrypted value leading to "IncompleteBody"
- Source is not encrypted and we provided source encryption keys.
This results in a corrupted object on the destination which is
considered encrypted but cannot be read by the server and returns
the following error.
```
<Error><Code>XMinioObjectTampered</Code><Message>The requested object
was modified and may be compromised</Message><Resource>/id-platform-gamma/
</Resource><RequestId>155EDC3E86BFD4DA</RequestId><HostId>3L137</HostId>
</Error>
```
CopyObject handler forgot to remove multipart encryption flag in metadata
when source is an encrypted multipart object and the target is also encrypted
but single part object.
This PR also simplifies the code to facilitate review.
This PR brings an additional logger implementation
called AuditLog which logs to http targets
The intention is to use AuditLog to log all incoming
requests, this is used as a mechanism by external log
collection entities for processing Minio requests.
Harshavardhana