That's odd because my own Entware setup is persistent and correctly survives reboots. I'll take another look, but make sure that you don't have leftovers from Optware still there.
For some reason, the '
ln -sf $1/entware.arm /tmp/opt' line in
post-mount wasn't overwriting an existing symbolic link to optware.
But on further investigation, there's something much stranger going on. '
disk monitor' is unmounting
/dev/sda1 right after it mounts, and isn't remounting it for almost 20 seconds. In the meantime,
services-start script runs, and the '
sleep 10' it starts with isn't nearly long enough for /dev/sda1 to have been remounted.
I also see a bunch of services being started well after
services-start was invoked (but while /dev/sda1 is still unmounted). I thought
services-start was supposed to be called after the system services had already been started?
Below I've pasted an excerpt from syslog from boot (I've replaced dnsmasq and dropbear 'noise' with '...'). My current
services-start script looks like this:
#!/bin/sh
logger $(basename $0)
log=/tmp/$(basename $0).log
sleep 60
mounted="`mount | grep '/dev/sda1'`"
logger $(basename $0) ${mounted}
/opt/etc/init.d/rc.unslung start > $log
This script 'works'; changing the '
sleep 60' to anything below 19 secs doesn't result in
rc.unslung being invoked. The default script uses '
sleep 10', so entware startup scripts are not being run.
Two things stand out to me:
a)
services-start seems to be invoked a lot earlier that I expected. A number of services haven't been started at that point.
b)
disk monitor unmounts
/dev/sda1 for an extended period during boot when I'd expect it would need to be mounted. Perhaps related to the kernel fsck message? if
disk monitor is doing an fsck, it's not resetting the flag, since it's doing it on every single boot.
I'd dig further, but spare time is short for the next few days. Let me know if there's anything I can do to help.
- Paul
---
Nov 30 16:00:15 kernel: sd 0:0:0:0: [sda] 15751168 512-byte logical blocks: (8.06 GB/7.51 GiB)
Nov 30 16:00:15 kernel: sd 0:0:0:0: [sda] Write Protect is off
Nov 30 16:00:15 kernel: sd 0:0:0:0: [sda] Assuming drive cache: write through
Nov 30 16:00:15 kernel: sd 0:0:0:0: [sda] Assuming drive cache: write through
Nov 30 16:00:15 kernel: sda: sda1
Nov 30 16:00:15 kernel: sd 0:0:0:0: [sda] Assuming drive cache: write through
Nov 30 16:00:15 kernel: sd 0:0:0:0: [sda] Attached SCSI removable disk
Nov 30 16:00:15 stop_nat_rules: apply the redirect_rules!
Nov 30 16:00:15 kernel: EXT2-fs (sda1): warning: mounting unchecked fs, running e2fsck is recommended
Nov 30 16:00:15 qcsapi: write qcsapi conf ok
Nov 30 16:00:15 WAN Connection: ISP's DHCP did not function properly.
Nov 30 16:00:15 hotplug[578]: USB ext2 fs at /dev/sda1 mounted on /tmp/mnt/sda1
Nov 30 16:00:16 disk monitor: unmount partition
Nov 30 16:00:17 syslog: USB partition unmounted from /tmp/mnt/sda1
Nov 30 16:00:17 disk monitor: scan partition
Nov 30 16:00:24 dropbear[668]: Running in background
Nov 30 16:00:27 dnsmasq[682]: started, version 2.73test6 cachesize 1500
Nov 30 16:00:27 dnsmasq[682]: warning: no addresses found for interface vlan2
Nov 30 16:00:27 dnsmasq[682]: warning: no addresses found for interface vlan2
Nov 30 16:00:27 dnsmasq[682]: warning: interface ppp1* does not currently exist
Nov 30 16:00:27 dnsmasq[682]: asynchronous logging enabled, queue limit is 5 messages
Nov 30 16:00:27 dnsmasq-dhcp[682]: DHCP, IP range 10.11.1.80 -- 10.11.1.99, lease time 1h
Nov 30 16:00:27 dnsmasq-dhcp[682]: DHCP, IP range 10.11.1.100 -- 10.11.1.199, lease time 1d
Nov 30 16:00:27 dnsmasq[682]: read /etc/hosts - 5 addresses
Nov 30 16:00:27 dnsmasq[682]: read /etc/hosts.dnsmasq - 1 addresses
Nov 30 16:00:27 dnsmasq-dhcp[682]: read /etc/ethers - 1 addresses
Nov 30 16:00:27 RT-AC87U: start httpd
Nov 30 16:00:27 crond[690]: crond: crond (busybox 1.20.2) started, log level 8
Nov 30 16:00:28 kernel: gro enabled with interval 2
Nov 30 16:00:28 disk monitor: be idle
Nov 30 16:00:28 syslog: Generating SSL certificate...
Nov 30 16:00:30 Samba Server: daemon is started
Nov 30 16:00:30 admin: services-start
Nov 30 16:00:31 dnsmasq[682]: read /etc/hosts - 5 addresses
Nov 30 16:00:31 dnsmasq[682]: read /etc/hosts.dnsmasq - 1 addresses
Nov 30 16:00:31 dnsmasq-dhcp[682]: read /etc/ethers - 1 addresses
Nov 30 16:00:31 admin: /jffs/scripts/dhcpc-event deconfig
Nov 30 16:00:31 rc_service: udhcpc 768:notify_rc start_firewall
Nov 30 16:00:31 dnsmasq[682]: read /etc/hosts - 5 addresses
Nov 30 16:00:31 dnsmasq[682]: read /etc/hosts.dnsmasq - 1 addresses
Nov 30 16:00:31 dnsmasq-dhcp[682]: read /etc/ethers - 1 addresses
Nov 30 16:00:31 dnsmasq[682]: using nameserver 75.75.76.76#53 for domain local
Nov 30 16:00:31 dnsmasq[682]: using nameserver 75.75.76.76#53 for domain bad
Nov 30 16:00:31 dnsmasq[682]: using nameserver 75.75.75.75#53 for domain local
Nov 30 16:00:31 dnsmasq[682]: using nameserver 75.75.75.75#53 for domain bad
Nov 30 16:00:31 dnsmasq[682]: using nameserver 75.75.75.75#53
Nov 30 16:00:31 dnsmasq[682]: using nameserver 75.75.76.76#53
Nov 30 16:00:31 admin: /jffs/scripts/dhcpc-event bound
Nov 30 16:00:31 wan: finish adding multi routes
Nov 30 16:00:31 rc_service: udhcpc 768:notify_rc stop_upnp
Nov 30 16:00:31 rc_service: waitting "start_firewall" via udhcpc ...
...
Nov 30 16:00:35 WAN Connection: WAN was restored.
Nov 30 16:00:35 start_nat_rules: apply the nat_rules(/tmp/nat_rules)!
...
Nov 30 16:00:48 disk monitor: re-mount partition
Nov 30 16:00:48 kernel: EXT2-fs (sda1): warning: mounting unchecked fs, running e2fsck is recommended
Nov 30 16:00:48 syslog: USB ext2 fs at /dev/sda1 mounted on /tmp/mnt/sda1
Nov 30 16:00:48 disk monitor: done
...
Nov 30 16:01:01 rc_service: skip the event: stop_upnp.
Nov 30 16:01:01 rc_service: udhcpc 768:notify_rc start_upnp
Nov 30 16:01:01 rc_service: waitting "start_firewall" via udhcpc ...
Nov 30 16:01:04 qtn: bootcfg.tgz exists
Nov 30 16:01:04 start_nat_rules: apply the nat_rules(/tmp/nat_rules_vlan2_vlan2)!
Nov 30 16:01:05 kernel: nf_conntrack_rtsp v0.6.21 loading
Nov 30 16:01:05 kernel: nf_nat_rtsp v0.6.21 loading
Nov 30 16:01:08 miniupnpd[1099]: HTTP listening on port 34181
Nov 30 16:01:08 miniupnpd[1099]: Listening for NAT-PMP/PCP traffic on port 5351
Nov 30 16:01:08 ddns update: ez-ipupdate: starting...
Nov 30 16:01:08 ddns update: connected to updates.dnsomatic.com (67.215.92.215) on port 80.
Nov 30 16:01:08 ddns update: request successful
Nov 30 16:01:08 ddns update: asusddns_update: 0
Nov 30 16:01:09 ddns: ddns update ok
Nov 30 16:01:09 dhcp client: bound 73.170.152.176 via 73.170.152.1 during 215985 seconds.
Nov 30 16:01:09 ntp: start NTP update
Mar 2 23:29:01 rc_service: hotplug 578:notify_rc restart_nasapps
Mar 2 23:29:01 rc_service: ntp 1132:notify_rc restart_upnp
Mar 2 23:29:01 rc_service: waitting "restart_nasapps" via hotplug ...
Mar 2 23:29:01 iTunes: daemon is stopped
Mar 2 23:29:01 FTP Server: daemon is stopped
Mar 2 23:29:02 Samba Server: smb daemon is stopped
Mar 2 23:29:02 kernel: gro disabled
Mar 2 23:29:02 Timemachine: daemon is stopped
Mar 2 23:29:02 kernel: gro enabled with interval 2
Mar 2 23:29:03 Samba Server: daemon is started
Mar 2 23:29:04 rc_service: ntp 1132:notify_rc restart_diskmon
Mar 2 23:29:04 rc_service: waitting "restart_upnp" via ntp ...
Mar 2 23:29:04 miniupnpd[1099]: shutting down MiniUPnPd
Mar 2 23:29:05 miniupnpd[1192]: HTTP listening on port 50541
Mar 2 23:29:05 miniupnpd[1192]: Listening for NAT-PMP/PCP traffic on port 5351
Mar 2 23:29:08 disk monitor: be idle
Mar 2 23:29:13 dfs: start dfs scan
Mar 2 23:29:13 nodfs_scan: complete
Mar 2 23:29:14 dfs: start dfs scan
Mar 2 23:29:14 nodfs_scan: complete
...
Mar 2 23:29:22 admin: services-start /dev/sda1 on /tmp/mnt/sda1 type ext2 (rw,nodev,relatime,errors=continue,user_xattr)
Mar 2 23:29:23 admin: Started asterisk from .
...
Mar 2 23:29:53 crond[690]: time disparity of 132928 minutes detected
Mar 2 23:29:53 rc_service: rc 1336:notify_rc restart_wrs
Mar 2 23:29:54 start_nat_rules: apply the nat_rules(/tmp/nat_rules_vlan2_vlan2)!