Hello Yehuda,

2014-04-04 9:35 GMT+02:00 Benedikt Fraunhofer
<[email protected]>:

> I'm currently trying to find a faster box with enough free space to
> reproduce that and capture logs.

Here's the complete log with "debug rgw 20" and "debug ms 1" of a
failed large-ish multipart upload.

As it's 71mb gzipped i placed it here:
 http://www.traced.net/PMhpycoXcb9OirMzGQfaHzoQaC/radosgw11.obfus.log.gz

2014-04-08 10:57:14.328232 7f37907d8700 10  uid=bf2.0 requested perm (type)=2,
 policy perm=2, user_perm_mask=2, acl perm=2
2014-04-08 10:57:14.328233 7f37907d8700  2 req 17480:0.000196:s3:
 POST /file-262144:complete_multipart:verifying op params
2014-04-08 10:57:14.328234 7f37907d8700  2 req 17480:0.000197:s3:
 POST /file-262144:complete_multipart:executing
2014-04-08 10:57:14.328284 7f37907d8700  2 req 17480:0.000247:s3:
 POST /file-262144:complete_multipart:http status=416
2014-04-08 10:57:14.328289 7f37907d8700  1 ====== req done
 req=0x7f37b80132d0 http_status=416 ======
2014-04-08 10:57:14.328290 7f37907d8700 20 process_request() returned -34

this is what s3cmd outputs:

file-262144 -> s3://7251b77d-ae36-449a-bb8a-99bd47f18947/file-262144
 [part 17476 of 17477, 15MB]
 15728640 of 15728640   100% in    0s    80.82 MB/s  done
file-262144 -> s3://7251b77d-ae36-449a-bb8a-99bd47f18947/file-262144
 [part 17477 of 17477, 4MB]
 4194304 of 4194304   100% in    0s    56.07 MB/s  done
ERROR: S3 error: 416 (InvalidRange):

It was a 256 gig file, the script does no magic and is rather silly:
dd 256gig from /dev/zero to a file and upload it using s3cmd.
========================
#! /bin/bash

set -x

SIZE=$((256*1024))
FN=file-${SIZE}
S3CMD=s3cmd
S3CMD=/home/bf/s3cmd-1.5.0-beta1/s3cmd
MULTIPART="--disable-multipart"
MULTIPART=""

BUCK="s3://$(uuidgen)/"
if [ ! -f $FN ]; then
        dd if=/dev/zero of=$FN bs=4M count=$(($SIZE/4)) &
fi
$S3CMD  mb $BUCK
wait

$S3CMD  put $MULTIPART $FN $BUCK
#sleep 1
#$S3CMD  get --force ${BUCK}${FN} file1.1
========================

Thx
 Benedikt
_______________________________________________
ceph-users mailing list
[email protected]
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Reply via email to