Hi!
* Removed custom version of memcached.
* Updated hardening rules.
* Fixed a case where the return error from a socket connection differred from
that of a TCP/IP socket.
http://launchpad.net/libmemcached/
Cheers,
-Brian
smime.p7s
Description: S/MIME cryptographic signature
Hey,
I'm trying to find out if Memcached allows to cache data from geo spread
sources.
Main goal is to use Memcached to cache data from distant database.
Below I put a simple diagram of my scenario.
Client usually connects to database which is in his region but sometimes he
use a lot of other dat
Comment #14 on issue 278 by lgromano...@gmail.com: Memcached 1.4.13
segfaults (error 4) when SASL enabled
http://code.google.com/p/memcached/issues/detail?id=278
Hi,
I'm observing similar segfault on memcached v1.4.15 (memcached 1.4.15-1)
from ArchLinux (stable), unfortunately I don't have
Hi,
I'm running memcached-1.4.14 in CentOS 6 x86_64 system, and my clients
write small objects in memcached. I'm staring memcached with this
directives:
memcached -d -p 11211 -u memcache -m 18432 -c 12288 -P
/var/run/memcached/memcached.pid -t 64
The memory has enough space, the slabs are not
free_chunks_end is just a counter for how many chunks are available in a
recently allocated slab page. So if a slab class grabs 1MB of new memory,
it'll have something in free_chunks_end temporarily, then it all moves
into free_chunks or otherwise gets used.
In 1.4.15 this counter is gone complete
Dormando,
Thank you for your reply!
I understood your explanation. But I installed 1.4.15 version in a test
server (not production) and the behaviour is the same:
# Item_Size Max_age Pages Count Full? Evicted Evict_Time OOM
2 224B59s 1 2 yes0
Maybe I forgot to remove the stat.
Again, you're not doing anything wrong. free_chunks_end is just a weird
indicator.
free_chunks == free_chunks + free_chunks_end. that's all.
On Wed, 19 Sep 2012, mmsilveira wrote:
> Dormando,
>
> Thank you for your reply!
>
> I understood your explanation. But
Actually I think I made that change in 1.4.14, not 1.4.15. but whatever :P
I'll remove the stat for the next cut, sorry!
On Wed, 19 Sep 2012, dormando wrote:
> Maybe I forgot to remove the stat.
>
> Again, you're not doing anything wrong. free_chunks_end is just a weird
> indicator.
>
> free_chun
No problem! :)
But, without free_chunks_end indicator, what about the column "Full?" of
memcached-tool?
Mauricio
Em quarta-feira, 19 de setembro de 2012 17h15min51s UTC-3, Dormando
escreveu:
>
> Actually I think I made that change in 1.4.14, not 1.4.15. but whatever :P
> I'll remove the stat
I think that's just misleading.
It means the slab is in full use... memcached can't actually tell if a
slab class is full or not. you measure that by eviction rate evicted_time,
evicted_nonfetched, etc. there's some discussion at the wiki.
On Wed, 19 Sep 2012, mmsilveira wrote:
> No problem! :)
Perfect.
One more time, thank you Dormando!
Mauricio
Em quarta-feira, 19 de setembro de 2012 17h28min16s UTC-3, Dormando
escreveu:
>
> I think that's just misleading.
>
> It means the slab is in full use... memcached can't actually tell if a
> slab class is full or not. you measure that by ev
11 matches
Mail list logo
