|
Index |
access control list (ACL), 1
ACL, using, 1
examples, 1
Lustre support, 1
configuring, 1
interpreting, 1
introduction, 1
clients, 1
OSTs, 1
allocating quotas, 1
bonding, 1
configuring Lustre, 1
module parameters, 1
references, 1
requirements, 1
setting up, 1
OSS memory requirements, 1
changelogs, 1
checksums, 1
eviction, 1
extents survey, 1
offset survey, 1
adding, 1
CollectL, 1
lfs, 1
lfsck, 1
mount, 1
Commit on Share (COS), 1
components, Lustre, 1
configuration example, Lustre, 1
configuration, logs, regenerating, 1
failover, 1
adaptive timeouts, 1
root squash, 1
debug_mb, 1
adding debugging to source code, 1
buffer, 1
controlling the kernel debug log, 1
finding Lustre UUID of an OST, 1
finding memory leaks, 1
looking at disk content, 1
messages, 1
Ptlrpc request history, 1
sample lctl run, 1
tracing lock traffic, 1
troubleshooting with strace, 1
directory statahead, using, 1
downed routers, 1
e2fsprogs, 1
e2scan, 1
Elan (Quadrics Elan), 1
environmental requirements, 1
error messages, 1
error numbers, 1
external journal, creating, 1
failover, 1
capabilities, 1
configuration types, 1
configuring, 1
failover and MMP, 1
MDT (active/passive), 1
OST (active/active), 1
file formats, quotas, 1
file readahead, using, 1
file size, maximum, 1
file striping, 1
file system size, maximum, 1
filename length, maximum, 1
flock utility, 1
adjusting weighting between free space and location, 1
round-robin allocator, 1
weighted allocator, 1
free space, managing, 1
HA software, 1
handling full OSTs, 1
handling timeouts, 1
checksums, 1
I/O tunables, 1
I/O, direct, performing, 1
inode number, OST, 1
inode size, MDT, 1
installing Lustre, environmental requirements, 1
installing Lustre, HA software, 1
installing Lustre, memory requirements, 1
installing Lustre, required tools / utilities, 1
interoperability, 1
adaptive timeouts, 1
key features, 1
lctl, 1
setting parameters, 1
lfs command, 1
lfsck command, 1
ll_recover_lost_found_objs, 1
llapi, 1
llstat, 1
starting, 1
stopping, 1
commands, 1
Load balancing with InfiniBand
modprobe.conf, 1
locking proc entries, 1
lst, 1
administration, aborting recovery, 1
administration, changing a server NID, 1
administration, determining which machine is serving an OST, 1
administration, failout / failover mode for OSTs, 1
administration, finding nodes in the file system, 1
administration, mounting a server, 1
administration, mounting a server without Lustre service, 1
administration, regenerating Lustre configuration logs, 1
administration, removing and restoring OSTs, 1
administration, running multiple Lustre file systems, 1
administration, setting and retrieving Lustre parameters, 1
administration, starting Lustre, 1
administration, working with inactive OSTs, 1
adminstration, unmounting a server, 1
components, 1
configuration example, 1
COS, 1
failover, 1
installing, environmental requirements, 1
installing, HA software, 1
installing, memory requirements, 1
installing, required tools / utilities, 1
interoperability, 1
key features, 1
metadata replay, 1
orphaned objects, 1
parameters, reporting current, 1
parameters, setting and retrieving, 1
parameters, setting with lctl, 1
parameters, setting with mkfs.lustre, 1
parameters, setting with tunefs.lustre, 1
recovering from corruption in Lustre file system, 1
recovering from errors on backing file system, 1
recovery, client eviction, 1
reply reconstruction, 1
scaling, 1
starting, 1
upgrading, 1.6.x to 1.8.x, 1
VBR, introduction, 1
VBR, tips, 1
VBR, working with, 1
downloading, 1
obdfilter_survey tool, 1
ost_survey tool, 1
sgpdd_survey tool, 1
Lustre Monitoring Tool (LMT), 1
lustre_rmmod.sh, 1
lustre_rsync, 1
lfs, 1
lfsck, 1
mount, 1
user/group cache upcall, 1
LNET options, 1
module options, 1
MX LND, 1
Portals LND (Linux), 1
SOCKLND kernel TCP/IP LND, 1
application profiling utilties, 1
lctl, 1
ll_recover_lost_found_objs, 1
llstat, 1
lst, 1
lustre_rsync, 1
mkfs.lustre, 1
mount.lustre, 1
plot-llstat, 1
proc statistics, 1
routerstat, 1
system configuration utilities, 1
test/debug utilities, 1
tunefs.lustre, 1
managing free space, 1
file size, 1
file system size, 1
filename/pathname length, 1
number of clients, 1
number of files in a directory, 1
number of open files, 1
number of OSTs and MDTs, 1
stripe count, 1
stripe size, 1
history, 1
tunables, 1
service thread count, 1
failover, 1
inode size, 1
memory requirements, 1
metadata replay, 1
stripe size, 1
mkfs.lustre, 1
setting parameters, 1
MMP and failover, 1
modprobe.conf, 1
changelogs, 1
CollectL, 1
Lustre Monitoring Tool, 1
mount command, 1
mount.lustre, 1
MX LND, 1
bonding, 1
Elan (Quadrics Elan), 1
ra (RapidArray), 1
NID, server, changing, 1
number of clients, maximum, 1
number of files in a directory, maximum, 1
number of open files, maximum, 1
obdfilter_survey tool, 1
orphaned objects, working with, 1
memory, determining, 1
service thread count, 1
OSS read cache, 1
failover, 1
number of inodes, 1
removing and restoring, 1
OST block I/O stream, watching, 1
OST pools, 1
OST, adding, 1
OST, determining which machine is serving, 1
ost_survey tool, 1
adding, 1
OSTs and MDTs, maximum, 1
OSTs, full, handling, 1
parameters, setting with lctl, 1
parameters, setting with mkfs.lustre, 1
parameters, setting with tunefs.lustre, 1
pathname length, maximum, 1
performing direct I/O, 1
Perl, 1
plot-llstat, 1
pools, OST, 1
Linux, 1
free space distribution, 1
LNET information, 1
locating file systems and servers, 1
locking, 1
timeouts, 1
Quadrics Elan, 1
quota limits, 1
quota statistics, 1
administering, 1
allocating, 1
creating files, 1
enabling, 1
file formats, 1
granted cache, 1
known issues, 1
limits, 1
statistics, 1
working with, 1
ra (RapidArray), 1
creating an external journal, 1
formatting options, 1
handling degraded arrays, 1
performance tradeoffs, 1
reliability best practices, 1
selecting storage for MDS or OSTs, 1
RapidArray, 1
readahead, tuning, 1
client failure, 1
MDS failure/failover, 1
network partition, 1
OST failure, 1
recovery, aborting, 1
regenerating configuration logs, 1
reply reconstruction, 1
reporting current Lustre parameters, 1
required tools / utilities, 1
configuring, 1
tips, 1
tuning, 1
root squash, using, 1
round-robin allocator, 1
routers, downed, 1
routerstat, 1
RPC stream tunables, 1
RPC stream, watching, 1
scaling Lustre, 1
mounting, 1
unmounting, 1
server NID, changing, 1
MDS, 1
OSS, 1
SCSI I/O sizes, 1
setting and retrieving Lustre parameters, 1
sgpdd_survey tool, 1
SOCKLND kernel TCP/IP LND, 1
LNET, 1
statahead, tuning, 1
LNET, 1
strace, 1
stripe count, maximum, 1
stripe size, maximum, 1
stripe size, minimum, 1
managing free space, 1
size, 1
striping using llapi, 1
Elan (Quadrics Elan), 1
ra (RapidArray), 1
timeouts, handling, 1
drawbacks in doing multi-client O_APPEND writes, 1
erasing a file system, 1
error messages, 1
error numbers, 1
handling timeouts on initial Lustre setup, 1
address already in use" error, 1
handling/debugging "Lustre Error
xxx went back in time", 1
handling/debugging error "28", 1
identifying a missing OST, 1
improving Lustre performance when working with small files, 1
log message ’out of memory’ on OST, 1
"slow start_page_write", 1
OST object missing or damaged, 1
OSTs become read-only, 1
reclaiming reserved disk space, 1
recovering from an unavailable OST, 1
replacing an existing OST or MDS, 1
reporting a Lustre bug, 1
setting SCSI I/O sizes, 1
slowdown occurs during Lustre startup, 1
triggering watchdog for PID NNN, 1
RPC stream, 1
tunables, lockless, 1
tunefs.lustre, 1
setting parameters, 1
directory statahead, 1
file readahead, 1
lockless tunables, 1
MDS threads, 1
OSS threads, 1
root squash, 1
1.6.x to 1.8.x, 1
complete file system, 1
e2fsprogs, 1
Perl, 1
VBR, introduction, 1
VBR, tips, 1
VBR, working with, 1
Version-based recovery (VBR), 1
weighted allocator, 1
weighting, adjusting between free space and location, 1
Copyright © 2011, Oracle and/or its affiliates. All rights reserved.