1
0
mirror of https://github.com/mariadb-corporation/mariadb-columnstore-engine.git synced 2025-07-05 15:41:14 +03:00

MCOL-4576: remove S3 options from cpimport. (#2307)

This commit is contained in:
benthompson15
2022-03-21 09:54:39 -05:00
committed by GitHub
parent 63c19a2413
commit 14c4840d53
3 changed files with 20 additions and 17 deletions

View File

@ -172,12 +172,11 @@ void printUsage()
<< " -T Timezone used for TIMESTAMP datatype" << endl
<< " Possible values: \"SYSTEM\" (default)" << endl
<< " : Offset in the form +/-HH:MM" << endl
<< endl
<< " -y S3 Authentication Key (for S3 imports)" << endl
<< " -K S3 Authentication Secret (for S3 imports)" << endl
<< " -t S3 Bucket (for S3 imports)" << endl
<< " -H S3 Hostname (for S3 imports, Amazon's S3 default)" << endl
<< " -g S3 Regions (for S3 imports)" << endl
// << " -y S3 Authentication Key (for S3 imports)" << endl
// << " -K S3 Authentication Secret (for S3 imports)" << endl
// << " -t S3 Bucket (for S3 imports)" << endl
// << " -H S3 Hostname (for S3 imports, Amazon's S3 default)" << endl
// << " -g S3 Regions (for S3 imports)" << endl
<< " -U username of new data files owner. Default is mysql" << endl;
cout << " Example1:" << endl
@ -310,7 +309,7 @@ void parseCmdLineArgs(int argc, char** argv, BulkLoad& curJob, std::string& sJob
BulkModeType bulkMode = BULK_MODE_LOCAL;
std::string jobUUID;
while ((option = getopt(argc, argv, "b:c:d:e:f:hij:kl:m:n:p:r:s:u:w:B:C:DE:I:P:R:ST:X:NL:y:K:t:H:g:U:")) !=
while ((option = getopt(argc, argv, "b:c:d:e:f:hij:kl:m:n:p:r:s:u:w:B:C:DE:I:P:R:ST:X:NL:U:")) !=
EOF)
{
switch (option)
@ -677,7 +676,7 @@ void parseCmdLineArgs(int argc, char** argv, BulkLoad& curJob, std::string& sJob
break;
}
case 'y':
/* case 'y':
{
curJob.setS3Key(optarg);
break;
@ -706,7 +705,7 @@ void parseCmdLineArgs(int argc, char** argv, BulkLoad& curJob, std::string& sJob
curJob.setS3Region(optarg);
break;
}
*/
case 'U':
{
curJob.setUsername(optarg);

View File

@ -561,11 +561,6 @@ void WECmdArgs::usage()
<< "\t-T\tTimezone used for TIMESTAMP datatype.\n"
<< "\t\tPossible values: \"SYSTEM\" (default)\n"
<< "\t\t : Offset in the form +/-HH:MM\n"
<< "\t-y\tS3 Authentication Key (for S3 imports)\n"
<< "\t-K\tS3 Authentication Secret (for S3 imports)\n"
<< "\t-t\tS3 Bucket (for S3 imports)\n"
<< "\t-H\tS3 Hostname (for S3 imports, Amazon's S3 default)\n"
<< "\t-g\tS3 Region (for S3 imports)\n"
<< "\t-L\tDirectory for the output .err and .bad files.\n"
<< "\t\tDefault is " << string(MCSLOGDIR);
@ -598,7 +593,7 @@ void WECmdArgs::parseCmdLineArgs(int argc, char** argv)
if (argc > 0)
fPrgmName = string(MCSBINDIR) + "/" + "cpimport.bin"; // argv[0] is splitter but we need cpimport
while ((aCh = getopt(argc, argv, "d:j:w:s:v:l:r:b:e:B:f:q:ihm:E:C:P:I:n:p:c:ST:Ny:K:t:H:g:U:L:")) != EOF)
while ((aCh = getopt(argc, argv, "d:j:w:s:v:l:r:b:e:B:f:q:ihm:E:C:P:I:n:p:c:ST:N:U:L:")) != EOF)
{
switch (aCh)
{
@ -906,7 +901,7 @@ void WECmdArgs::parseCmdLineArgs(int argc, char** argv)
fConsoleOutput = false;
break;
}
/*
case 'y': //-y S3 Key
{
fS3Key = optarg;
@ -936,7 +931,7 @@ void WECmdArgs::parseCmdLineArgs(int argc, char** argv)
fS3Region = optarg;
break;
}
*/
case 'U': //-U username of the files owner
{
fUsername = optarg;

View File

@ -481,6 +481,15 @@ void WEFileReadThread::openInFile()
use ms3 lib to d/l data into mem
use boost::iostreams to wrap the mem in a stream interface
point infile's stream buffer to it.
MCOL-4576: The options to setup S3 with cpimport have been removed and this
code is unreachable. However we may need to resurrect it at some point in some form.
Performance issues with extremely large data files as well as the fact files larger
than system memory will cause an OOM error. Multipart downloads/uploads need to be
implemented or more likely a different streaming solution developed with external API tools
MCOL-4576 work around is to use 3rd party CLI tools and pipe data file from S3 bucket
into cpimport stdin. 3rd party tooling for large object downloads will be more efficient.
*/
if (fSdh.getDebugLvl())