On 11/15/2017 11:40 AM, Metztli Information Technology wrote:
Niltze, Ed-
Here is some additional feedback on reiser4 transparent-compression:
Under Google Compute Engine (GCE) using resources: --custom-cpu=2 --custom-memory=4GB
Below dmesg lapse occurred while using VBoxManage (command line utility from headless VirtualBox 5.1.30) under latest reiser4-for-4.13.1.patch kernel
< https://sourceforge.net/projects/metztli-reiser4/files/Reiser4-Linux-4.13.12-1_for-Stretch/ >
while converting a 2TB virtual machine into its disk.raw equivalent (11-14-2017):
[..]
[ 207.416756] reiser4: sda4: found disk format 4.0.1.
[ 207.488274] reiser4: sda4: using Hybrid Transaction Model.
[ 965.372697] sdb: sdb1
[ 1020.422370] reiser4: sdb1: found disk format 4.0.1.
[ 1044.098194] reiser4: sdb1: using Hybrid Transaction Model.
[ 2256.307429] vboxdrv: loading out-of-tree module taints kernel.
[ 2256.314916] vboxdrv: Found 2 processor cores
[ 2256.333498] vboxdrv: TSC mode is Invariant, tentative frequency 2199993485 Hz
[ 2256.340792] vboxdrv: Successfully loaded version 5.1.30 (interface 0x002a0000)
[ 2265.565041] VBoxNetFlt: Successfully started.
[ 4898.939952] reiser4[ktxnmgrd:sdb1:r(12134)]: commit_current_atom (/usr/src/build/kernel/tekitl-4.13.12/linux/fs/reiser4/txnmgr.c:1052)[nikita-317
6]:
WARNING: Flushing like mad: 16384
Along with complaint "Too many iterations" it is to detect silent
corruptions of a storage tree. If it is rare, then no problems.
If it is flood-like, then "hello fsck".
------------------------------------------------
Older:
with previous patch reiser4-for-4.13.0.patch.gz: logged neary 5k lines snapshot
during heavy load (using VirtualBox 5.1.28/30, etc.) and other applications in my local machine.
Had to reboot machine as there was no way to stop dmesg flow (Oct 22, 2017):
[24068.404896] reiser4[DOMCacheThread(29961)]: assign_conversion_mode (/usr/src/build/kernel/tekitl-4.13/linux/fs/reiser4/plugin/item/ctail.c:1588)[edward-1021]:
WARNING: convert_ctail: nothing to attach
[24068.404938] reiser4[DOMCacheThread(29961)]: assign_conversion_mode (/usr/src/build/kernel/tekitl-4.13/linux/fs/reiser4/plugin/item/ctail.c:1588)[edward-1021]:
WARNING: convert_ctail: nothing to attach
[24068.404950] reiser4[DOMCacheThread(29961)]: assign_conversion_mode (/usr/src/build/kernel/tekitl-4.13/linux/fs/reiser4/plugin/item/ctail.c:1588)[edward-1021]:
WARNING: convert_ctail: nothing to attach
[24068.404962] reiser4[DOMCacheThread(29961)]: assign_conversion_mode (/usr/src/build/kernel/tekitl-4.13/linux/fs/reiser4/plugin/item/ctail.c:1588)[edward-1021]:
WARNING: convert_ctail: nothing to attach
[24068.404974] reiser4[DOMCacheThread(29961)]: assign_conversion_mode (/usr/src/build/kernel/tekitl-4.13/linux/fs/reiser4/plugin/item/ctail.c:1588)[edward-1021]:
WARNING: convert_ctail: nothing to attach
[24068.404986] reiser4[DOMCacheThread(29961)]: checkout_logical_cluster (/usr/src/build/kernel/tekitl-4.13/linux/fs/reiser4/plugin/file/cryptcompress.c:1980)[edward-1430]:
WARNING: alloc stream failed with ret=-12
Here kernel can not allocate 64K for compression. This is due to
a "progress" in kernel's memory management. Ugly workarounds are
required. We need to teach the flush procedure to insert uncompressed
logical chunk represented by a number of dirty pages.
Edward.
--
To unsubscribe from this list: send the line "unsubscribe reiserfs-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
