Jump to content
XPEnology Community

snakefox666

Member
 View Profile  See their activity

  • Posts

    37

  • Joined

  • Last visited

 Content Type 

Posts posted by snakefox666

    Memory usage are increasing over weeks and the NAS finally crashed

    in General Questions

    Posted

    Hello,

     

    I've a strange issue with my brand new DS 916+, I had previously a HP N40L with xpenology and after more than 5 years of good services I decided to buy a real synology :)

     

    The RAM usage of my DS916+ (8Gb) is increasing over weeks and finally the nas is crashing. After couples of investigations it seem to be related to slab memory usage that increasing dramatically. Bellow you will find the result of some commands related to slab functionnality. We can see that "cred_jar" is using most of the memory and it's "Unreclaimable" :-(

    After a reboot of the nas the memory start at more or less 20% of usage and is incresing over 2 or 3 weeks to more than 90%.


      

    ash-4.4# cat /proc/meminfo
MemTotal: 8093200 kB
MemFree: 1224388 kB
Buffers: 9892 kB
Cached: 948788 kB
SwapCached: 42332 kB
Active: 652760 kB
Inactive: 1078964 kB
Active(anon): 384036 kB
Inactive(anon): 885104 kB
Active(file): 268724 kB
Inactive(file): 193860 kB
Unevictable: 1184 kB
Mlocked: 1184 kB
SwapTotal: 6954924 kB
SwapFree: 6654860 kB
Dirty: 136 kB
Writeback: 0 kB
AnonPages: 742560 kB
Mapped: 312764 kB
Shmem: 494924 kB
Slab: 4547088 kB
SReclaimable: 77560 kB
SUnreclaim: 4469528 kB
KernelStack: 13104 kB
PageTables: 60516 kB
NFS_Unstable: 0 kB
Bounce: 0 kB
WritebackTmp: 0 kB
CommitLimit: 11001524 kB
Committed_AS: 7307152 kB
VmallocTotal: 34359738367 kB
VmallocUsed: 525616 kB
VmallocChunk: 34359113488 kB
DirectMap4k: 19200 kB
DirectMap2M: 8282112 kB


ash-4.4# slabtop --sort c
Active / Total Objects (% used) : 37272947 / 37437940 (99.6%)
Active / Total Slabs (% used) : 1131412 / 1131432 (100.0%)
Active / Total Caches (% used) : 147 / 269 (54.6%)
Active / Total Size (% used) : 4192989.21K / 4223942.65K (99.3%)
Minimum / Average / Maximum Object : 0.02K / 0.11K / 4096.00K

OBJS ACTIVE USE OBJ SIZE SLABS OBJ/SLAB CACHE SIZE NAME
18127720 18127720 100% 0.19K 906386 20 3625544K cred_jar
18155200 18155200 100% 0.03K 162100 112 648400K kmalloc-32
494020 493871 99% 0.19K 24701 20 98804K kmalloc-192
30919 11710 37% 0.55K 4417 7 17668K radix_tree_node
25398 24524 96% 0.63K 4233 6 16932K inode_cache
64008 61953 96% 0.18K 3048 21 12192K vm_area_struct
57920 40462 69% 0.19K 2896 20 11584K dentry
670 655 97% 16.00K 670 1 10720K kmalloc-16384
272 272 100% 32.00K 272 1 8704K kmalloc-32768
2133 2133 100% 4.00K 2133 1 8532K kmalloc-4096
107484 103436 96% 0.07K 2028 53 8112K btrfs_free_space
864 862 99% 8.00K 864 1 6912K kmalloc-8192
4098 4096 99% 1.20K 1366 3 5464K raid5-md2
3768 2229 59% 1.19K 1256 3 5024K btrfs_inode
53900 10859 20% 0.09K 1225 44 4900K btrfs_extent_state
70092 60033 85% 0.06K 1188 59 4752K anon_vma_chain
5705 5676 99% 0.70K 1141 5 4564K proc_inode_cache
14745 10194 69% 0.25K 983 15 3932K filp
3368 3311 98% 1.00K 842 4 3368K kmalloc-1024
4055 4020 99% 0.70K 811 5 3244K shmem_inode_cache
46551 35428 76% 0.06K 789 59 3156K kmalloc-64
2244 1915 85% 1.02K 748 3 2992K ext4_inode_cache
43282 32811 75% 0.05K 646 67 2584K anon_vma
5096 5096 100% 0.50K 637 8 2548K kmalloc-512
8151 1422 17% 0.30K 627 13 2508K btrfs_delayed_node
18060 6824 37% 0.12K 602 30 2408K kmalloc-128
20094 20072 99% 0.11K 591 34 2364K sysfs_dir_cache
962 962 100% 2.00K 481 2 1924K kmalloc-2048
880 877 99% 1.97K 440 2 1760K task_struct
5605 1205 21% 0.20K 295 19 1180K btrfs_extent_buffer
9 9 100% 128.00K 9 1 1152K kmalloc-131072
405 385 95% 2.06K 135 3 1080K sighand_cache
7440 3999 53% 0.12K 248 30 992K kmalloc-96
273 256 93% 2.06K 91 3 728K idr_layer_cache
860 811 94% 0.69K 172 5 688K sock_inode_cache
5550 2592 46% 0.10K 150 37 600K buffer_head


ash-4.4# cat /proc/slabinfo | grep -i cred_jar
cred_jar 18127976 18128000 192 20 1 : tunables 120 60 8 : slabdata 906400 906400 0

     

    I tried to contact the synology support but the only answer is "Grant us remote control" and I don't want to let them access my private NAS :(

    Any help is welcome, thanks guys 😎

     

     

    DSM 6.2 Loader

    in Loaders

    Posted

    6 hours ago, synofreak said:

    Hi @Jun THX for great Update! :10_wink:

     

    HP Microserver N40L no Problems at the Moment!!!

     

    cu-ya

     

    Did you update from previous version or it's a fresh install ? I'm curious because I have an N40L too ;)

    Thanks mate !

    DSM 6.1.x Loader

    in Loaders

    Posted

    I don't know if it can help for the UPS issue on HP microserver :

     

    #4: OHCI support

     

    Current releases of the XPEnology lacks OHCI USB support (remember: USB 1.1 hardware has two different driver models, UHCI and OHCI; also all USB 2.0 hosts has two drivers, one for USB 2.0 protocol (EHCI) and one for USB 1.1 (UHCI or OHCI)). This is the reason because in the HP Microserver you lack USB support for low speed devices (like keyboard, mouse, UPS, etc.). The kernel loads the correct EHCI driver, but for USB 1.1 speeds only has the UHCI driver and the system can't comunicate with your devices.

     

    Here the solution: Add support for OHCI driver.

    You only need to modify the config file in the kernel with:

     

    CONFIG_USB_ARCH_HAS_OHCI=y
CONFIG_USB_OHCI_HCD=m
CONFIG_USB_OHCI_HCD_SSB=y

     

    And this will generate the correct "ohci-hcd.ko" module. This module will be automatically loaded at boot, so you only need to copy them to "/lib/modules/".

     

    WARNING:

     

    After recompiling this module the module "usbcore.ko" is different. If you try to load the "ohci-hcd.ko" in a system with the original "usbcore.ko" module already loaded you receive a kernel fault. You need to replace also the "usbcore.ko" module!!! :eek:

     

    FOR MICROSERVER USERS:

     

    If you like to complete your USB support in your machine:

     

    1) Download the extended package of kernel modules (modules-4.2-02.05.2013.tar.gz ) and copy all of them in the directory "/lib/modules/": https://docs.google.com/file/d/0B5tpWhE ... _web&pli=1

    2) Download the new "usbcore.ko" and "ohci-hcd.ko" modules and copy them to "/lib/modules/": http://pastebin.com/sH4RieEj / http://pastebin.com/4vL65THx

    3) Reboot and you have full USB support (all ports, USB 1.1 and USB 2.0 speeds)!

     

    :grin:

     

    I want to get my USB external hard drive working with my HP microserver.

     

    I have a few questions:

    1) do I need to do both steps? Ie add support for OHCI driver and copy across the kernel modules?

    2) if so, which file do I need to modify in the kernel and how do you do this?

    3) what's the best method for copying across the modules?

     

    It would be great if someone could post a step by step guide to doing this?

     

    Thanks

    Mike

     

    From viewtopic.php?f=2&t=915&start=10

    DSM 6.1.x Loader

    in Loaders

    Posted

    hello with the last loader on my N40L, the RTL8153 in my hub usb3 doesn't work :sad:

    I found why my UPS was no longer connected, the USB chipset original for N40L does not work, I plugged in my USB3 hub in my PCI usb3 and is now detected

     

    Hey jun, do you think that it's possible to integrate the driver for USB chipset of the N40L into your image ? :roll:

     

    +1

     

    Just a tiny up :roll:

    If someone want to help on that and need more information don't hesitate to ask ...

    [Tuto] Installer/Migrer DSM 5.2 à 6.0.2 (Loader Jun)

    in Archives

    Posted

    Bonjour,

    Comment savez-vous que cela fonctionne car sur mon N54L en lançant la commande ci-dessous dans putty mon processeur reste à sa fréquence max de 2200Mhh.

    grep "MHz" /proc/cpuinfo

    Mais en créant un script utilisateur au démarrage avec l'utilisateur root j'ai mon processeur qui descend à la fréquence de 800Mhz.

    code à mettre

    for c in $(ls -d /sys/devices/system/cpu/cpu[0-9]*);
do if ! grep -q 'conservative' $c/cpufreq/scaling_governor ;
   then echo conservative >$c/cpufreq/scaling_governor ;
fi
done

    Et là l'économie d'énergie est activée.

     

    C'était sous entendu en fait qu'il faut utiliser le script ci-dessus, en effet de base après l'installation le CPU est au max. Désolé si ce n'était pas claire.

    [Tuto] Installer/Migrer DSM 5.2 à 6.0.2 (Loader Jun)

    in Archives

    Posted

    Il existe même un script pour gérer l'économie d'énergie (S99PowersavingAMD.sh).

    #!/bin/sh

insmod /lib/modules/powernow-k8.ko

for c in $(ls -d /sys/devices/system/cpu/cpu[0-9]*);
do if ! grep -q 'conservative' $c/cpufreq/scaling_governor ;
   then echo conservative >$c/cpufreq/scaling_governor ;
fi
done

     

    Bàv

    Youpie le portage a été fait et fonctionne très bien.

    Savez-vous si l'économie d’énergie est intégrée?

    Et si un support pour les UPS est prévu?

     

    Je ne sais pas si quelqu'un à répondu mais je confirme que sur un N40L la fonction d'économie d’énergie marche impeccable :cool:

    DSM 6.1.x Loader

    in Loaders

    Posted

    hello with the last loader on my N40L, the RTL8153 in my hub usb3 doesn't work :sad:

    I found why my UPS was no longer connected, the USB chipset original for N40L does not work, I plugged in my USB3 hub in my PCI usb3 and is now detected

     

    Hey jun, do you think that it's possible to integrate the driver for USB chipset of the N40L into your image ? :roll:

    DSM 6.1.x Loader

    in Loaders

    Posted

    Hello All,

    I have updated my HP N40L successfully, thanks for the authors of this loader !!!

    Just a "little" thing regarding the UPS, nothing was detected and it was working before in 5.2 (USB UPS) :???:

    Any idea ?

    Strange crash

    in Archives

    Posted

    Hello all,

     

    Since yesterday I have a very strange issue with my Xpenology (HP N40L). Everytime I do a "ls" on the / directory all the services of my NAS was gone and not working, Samba, Syno webpage, all ... I can reboot the NAS and everything is back again until I connect in SSH and do "ls" in / again :shock:

     

    There is nothing special in the logs, it seem that when the thing happen the system create a new folder for /var and /run :

     

    mini_94305920160602223242PuTTYinactive.jpg

     

    for example if I'm doing a ls on this new /var or /run folder I can find "new" files since my crashing command. I can't enter the 2 strange folders with very strange characters "no such files or directory" :???:

     

    I already tried to unmount the partition and do fsck but the same...

     

    That's very very strange, that's not a big deal as for the moment I don't need to use the SSH :lol:

    Any idea is welcome :lol:

×
×
  • Create New...