The jffs partition is full

[Chiny91]

I now have similar on my RT-AC86U; I can't delete files, example:

/jffs# rm -f syslog.log-1
rm: can't remove 'syslog.log-1': No space left on device

jffs is not full:

df -h
Filesystem Size Used Available Use% Mounted on
[other stuff]
/dev/mtdblock9 47.0M 29.4M 17.6M 63% /jffs

After a power off reboot, it looks like a re-format

 

[OBENZ]

I’ve had the same problem for a while and I can’t even remove stuff as above I restored to default but still same issue after restoring my saved settings

Apr  5 12:28:01 watchdog: start ddns.
Apr  5 12:28:01 rc_service: watchdog 1083:notify_rc start_ddns
Apr  5 12:28:01 kernel: jffs2: Argh. No free space left for GC. nr_erasing_blocks is 0. nr_free_blocks is 0. (erasableempty: yes, erasingempty: yes, erasependingempty: yes)
Apr  5 12:28:01 kernel: jffs2: jffs2_reserve_space_gc of 196 bytes for garbage_collect_dnode failed: -28
Apr  5 12:28:01 kernel: jffs2: Error garbage collecting node at 02d76b8c!
Apr  5 12:28:19 kernel: [BLOCKED - INBOUND] IN=ppp0 OUT= MAC= SRC=74.120.14.93 DST=160.179.22.237 LEN=44 TOS=0x08 PREC=0x40 TTL=36 ID=5229 PROTO=TCP SPT=6180 DPT=18000 SEQ=3102300340 ACK=0 WINDOW=1024 RES=0x00 SYN URGP=0 OPT (020405B4) MARK=0x8000000
Apr  5 12:28:28 kernel: [BLOCKED - INBOUND] IN=ppp0 OUT= MAC= SRC=5.253.84.226 DST=160.179.22.237 LEN=40 TOS=0x08 PREC=0x40 TTL=241 ID=54321 PROTO=TCP SPT=52990 DPT=5909 SEQ=3830981291 ACK=0 WINDOW=65535 RES=0x00 SYN URGP=0 MARK=0x8000000
Apr  5 12:28:31 watchdog: start ddns.
Apr  5 12:28:31 rc_service: watchdog 1083:notify_rc start_ddns
Apr  5 12:28:31 custom_script: Running /jffs/scripts/service-event (args:  1)
Apr  5 12:28:31 custom_script: Running /jffs/scripts/service-event-end (args:  1)
Apr  5 12:28:56 kernel: [BLOCKED - INBOUND] IN=ppp0 OUT= MAC= SRC=162.142.125.27 DST=160.179.22.237 LEN=44 TOS=0x08 PREC=0x40 TTL=35 ID=56651 PROTO=TCP SPT=41078 DPT=1194 SEQ=2870026385 ACK=0 WINDOW=1024 RES=0x00 SYN URGP=0 OPT (020405B4) MARK=0x8000000
Apr  5 12:29:01 watchdog: start ddns.
Apr  5 12:29:01 rc_service: watchdog 1083:notify_rc start_ddns
Apr  5 12:29:01 kernel: jffs2: Argh. No free space left for GC. nr_erasing_blocks is 0. nr_free_blocks is 0. (erasableempty: yes, erasingempty: yes, erasependingempty: yes)
Apr  5 12:29:01 kernel: jffs2: jffs2_reserve_space_gc of 196 bytes for garbage_collect_dnode failed: -28
Apr  5 12:29:01 kernel: jffs2: Error garbage collecting node at 02d76b8c!

 

[Chiny91]

I’ve had the same problem for a while and I can’t even remove stuff as above I restored to default but still same issue after restoring my saved settings

I had that problem, so after (twice) reformatting jffs and restoring my saved settings to no avail, I reformatted jffs (for the third time) and re-installed various items from scratch, not restoring from a saved backup file. It now works fine. Presumably the saved backup also backed up something nasty, perhaps a corrupted file.

 

[OBENZ]

I tried again started from scratch but I keep getting the same error. I reset everything then started customising but the jffs error keep is popping up I haven’t even installed any custom scripts yet

 

[OBENZ]

Even when setting the location to ram I get the same error. Also when trying to update the firmware manually I upload the file then it says 100% complete but nothing happens the router doesn’t reboot and I’m still on the old fw

 

[gattaca]

Per -> Apr 5 12:28:01 kernel: jffs2: Error garbage collecting node at 02d76b8c!

There's likely a bad NAND block which may be preventing the GC routines from working. This kicked my a$$ around the block and back with 384.18/19 on an AC86U. I'm sure this is discussed in that long thread about what we were seeing.

IMHO, a @L&LD reset may be coming but you can also try this to save anything in the /JFFS:.

a) See if you can take a backup of the JFFS in GUI to save whatever. You may have to resort to copying files off individually as well.
b) Turn off the JFFS, reboot
c) Verify you have the latest firmware updates installed - which fixed a bad block = no format scenario, if not, install, reboot the router properly.
d) Turn on JFFS
e) reboot the router
f) Monitor the log closely.. you will see the format when booting and likely it will ID the bad block.

Longer discussion in here -> https://www.snbforums.com/threads/asuswrt-merlin-384-19-is-now-available.65801/page-24#post-615444

The root problem is the NAND garbage collection needs larger contiguous free-space to do reclaim/compaction/trim... and when there's a bad block, that can cause that contiguous space to be too small. If the format jffs on boot cannot "map around" the bad blocks during the format, then the problem will persist. The bad block will be added to a bad-block table by the format if all goes well. You may have to turn on "debug" in the logs to see those messages. Good Luck! Stay safe, stay alive! Peace.

 

[heysoundude]

I had that problem, so after (twice) reformatting jffs and restoring my saved settings to no avail, I reformatted jffs (for the third time) and re-installed various items from scratch, not restoring from a saved backup file. It now works fine. Presumably the saved backup also backed up something nasty, perhaps a corrupted file.

exactly right - the problem was restoring saved settings. starting from scratch might seem like a daunting task, but it often is necessary to clean the slate for better performance.
#ThisIsTheWay

 

[OBENZ]

Per -> Apr 5 12:28:01 kernel: jffs2: Error garbage collecting node at 02d76b8c!

There's likely a bad NAND block which may be preventing the GC routines from working. This kicked my a$$ around the block and back with 384.18/19 on an AC86U. I'm sure this is discussed in that long thread about what we were seeing.

IMHO, a @L&LD reset may be coming but you can also try this to save anything in the /JFFS:.

a) See if you can take a backup of the JFFS in GUI to save whatever. You may have to resort to copying files off individually as well.
b) Turn off the JFFS, reboot
c) Verify you have the latest firmware updates installed - which fixed a bad block = no format scenario, if not, install, reboot the router properly.
d) Turn on JFFS
e) reboot the router
f) Monitor the log closely.. you will see the format when booting and likely it will ID the bad block.

Longer discussion in here -> https://www.snbforums.com/threads/asuswrt-merlin-384-19-is-now-available.65801/page-24#post-615444

The root problem is the NAND garbage collection needs larger contiguous free-space to do reclaim/compaction/trim... and when there's a bad block, that can cause that contiguous space to be too small. If the format jffs on boot cannot "map around" the bad blocks during the format, then the problem will persist. The bad block will be added to a bad-block table by the format if all goes well. You may have to turn on "debug" in the logs to see those messages. Good Luck! Stay safe, stay alive! Peace.

Thanks for the reply but after nuking the router and starting from scratch my jffs partition shows only 1mb used now. It toi some time to get everything as it was but the end result is worth it.is there anyway to avoid this from happening again?i know @merlin says the traffic analyser is mostly the cause when it fails to flash the history db after reaching certain history/ size

 

[gattaca]

I stopped using the Traffic History DB a year or so ago.. just too many problems to leave running 24/7. Glad you got things resolved!