[CRIU] Re-Checkpoint Applications
Pavel Emelyanov
xemul at virtuozzo.com
Thu Feb 25 02:42:00 PST 2016
On 02/23/2016 07:09 AM, Abdullah Yousafzai wrote:
>
> script:
>
>
> while :; do
> sleep 1
> date
> done
>
> commands:
>
> setsid ./test.sh < /dev/null &> test.log &
>
> criu dump -t 2221 -vvv -o dump.log
>
> criu restore -d -vvv -o restore.log
Here you should use the --restore-detached option. It will make criu exit
after restore and re-parent your app to init process.
Additionally, you can use the --pidfile <filename> option and criu would
print the restore app's pid into the file named <filename> :)
-- Pavel
> but after that when I:
>
> ps -C test.sh
>
> I did not find the script, but when I did :
>
> ps -C criu
>
> the i found it and when I want to kill this process I am unable to kill with:
>
> kill criu_PID
>
> and the test.log is still updating the date. But if checkpoint the criu process shown in the ps command it fails and gives me the following log message:
>
> (00.000022) Probing sock diag modules
> (00.000094) Done probing
> (00.000100) ========================================
> (00.000107) Dumping processes (pid: 2672)
> (00.000110) ========================================
> (00.000143) Pagemap is fully functional
> (00.000181) Found anon-shmem device at 5
> (00.000188) Reset 27061's dirty tracking
> (00.000211) ... done
> (00.006545) Dirty tracking support is OFF
> (00.006609) Error (proc_parse.c:839): Can't open 27061/loginuid on procfs: No such file or directory
> (00.006623) irmap: Searching irmap cache in work dir
> (00.006629) No irmap-cache image
> (00.006630) irmap: Searching irmap cache in parent
> (00.006632) irmap: No irmap cache
> (00.006699) Perparing image inventory (version 1)
> (00.006727) Add pid ns 1 pid 27061
> (00.006737) Add net ns 2 pid 27061
> (00.006743) Add ipc ns 3 pid 27061
> (00.006747) Add uts ns 4 pid 27061
> (00.006751) Add mnt ns 5 pid 27061
> (00.006754) Add user ns 6 pid 27061
> (00.006755) cg: Dumping cgroups for 27061
> (00.006769) cg: Set 1 is criu one
> (00.006819) Seized task 2672, state 1
> (00.006831) Collected (4 attempts, 0 in_progress)
> (00.006840) Seized task 2640, state 0
> (00.006881) Collected (4 attempts, 0 in_progress)
> (00.006888) Collected (4 attempts, 0 in_progress)
> (00.006916) Collected 2640 in 1 state
> (00.006939) Collected (3 attempts, 0 in_progress)
> (00.006941) Collected 2672 in 1 state
> (00.006995) Lock network
> (00.007016) lockinfo: 1:1 0 20461 00:0e:8656 128 128
> (00.007032) lockinfo: 2:1 0 20461 00:0e:8653 1073741826 1073742335
> (00.007035) lockinfo: 3:1 0 15462 00:0e:156213 128 128
> (00.007038) lockinfo: 4:1 0 15462 00:0e:156210 1073741826 1073742335
> (00.007040) lockinfo: 5:1 1 15462 00:0e:156192 1073741824 1073742335
> (00.007042) lockinfo: 6:2 1 1620 00:0e:7867 0 EOF
> (00.007045) lockinfo: 7:1 0 1325 00:0e:7733 128 128
> (00.007047) lockinfo: 8:1 0 1325 00:0e:7730 1073741826 1073742335
> (00.007100) type tmpfs source tmpfs mnt_id 13 s_dev 0xd / @ ./ flags 0x300001 options
> (00.007108) type ext4 source /dev/loop1 mnt_id 16 s_dev 0x700001 / @ ./system flags 0x300001 options data=ordered
> (00.007112) type tmpfs source tmpfs mnt_id 17 s_dev 0xe / @ ./data flags 0x300000 options
> (00.007116) type tmpfs source tmpfs mnt_id 18 s_dev 0xf / @ ./dev flags 0x300002 options mode=755
> (00.007120) type devpts source devpts mnt_id 19 s_dev 0xb / @ ./dev/pts flags 0x300000 options mode=600
> (00.007128) type proc source proc mnt_id 20 s_dev 0x4 / @ ./proc flags 0x300000 options
> (00.007131) type sysfs source sysfs mnt_id 21 s_dev 0xc / @ ./sys flags 0x300000 options
> (00.007135) type debugfs source debugfs mnt_id 22 s_dev 0x6 / @ ./sys/kernel/debug flags 0x300000 options
> (00.007138) type cgroup source none mnt_id 23 s_dev 0x10 / @ ./acct flags 0x300000 options cpuacct
> (00.007142) type tmpfs source none mnt_id 24 s_dev 0x11 / @ ./sys/fs/cgroup flags 0x300000 options mode=750,gid=1000
> (00.007146) type tmpfs source tmpfs mnt_id 25 s_dev 0x12 / @ ./mnt/asec flags 0x300000 options mode=755,gid=1000
> (00.007150) type tmpfs source tmpfs mnt_id 26 s_dev 0x13 / @ ./mnt/obb flags 0x300000 options mode=755,gid=1000
> (00.007153) type cgroup source none mnt_id 27 s_dev 0x14 / @ ./dev/cpuctl flags 0x300000 options cpu
> (00.007156) type pstore source pstore mnt_id 28 s_dev 0x15 / @ ./sys/fs/pstore flags 0x300000 options
> (00.007160) type tmpfs source none mnt_id 29 s_dev 0x16 / @ ./cache flags 0x100406 options
> (00.007167) type fuse source /dev/fuse mnt_id 30 s_dev 0x17 / @ ./mnt/shell/emulated flags 0x30000e options user_id=1023,group_id=1023,default_permissions,allow_other
> (00.007173) mnt: Building mountpoints tree
> (00.007176) mnt: Resorting siblings on 13
> (00.007177) mnt: Resorting siblings on 30
> (00.007178) mnt: Resorting siblings on 29
> (00.007178) mnt: Resorting siblings on 26
> (00.007179) mnt: Resorting siblings on 25
> (00.007180) mnt: Resorting siblings on 23
> (00.007180) mnt: Resorting siblings on 21
> (00.007181) mnt: Resorting siblings on 28
> (00.007182) mnt: Resorting siblings on 24
> (00.007182) mnt: Resorting siblings on 22
> (00.007183) mnt: Resorting siblings on 20
> (00.007184) mnt: Resorting siblings on 18
> (00.007186) mnt: Resorting siblings on 27
> (00.007187) mnt: Resorting siblings on 19
> (00.007187) mnt: Resorting siblings on 17
> (00.007188) mnt: Resorting siblings on 16
> (00.007189) mnt: Done:
> (00.007189) mnt: [./](13->0)
> (00.007190) mnt: [./cache](29->13)
> (00.007191) mnt: <--
> (00.007192) mnt: [./system](16->13)
> (00.007193) mnt: <--
> (00.007194) mnt: [./data](17->13)
> (00.007194) mnt: <--
> (00.007195) mnt: [./dev](18->13)
> (00.007196) mnt: [./dev/cpuctl](27->18)
> (00.007197) mnt: <--
> (00.007197) mnt: [./dev/pts](19->18)
> (00.007198) mnt: <--
> (00.007199) mnt: <--
> (00.007199) mnt: [./proc](20->13)
> (00.007200) mnt: <--
> (00.007201) mnt: [./sys](21->13)
> (00.007202) mnt: [./sys/fs/pstore](28->21)
> (00.007203) mnt: <--
> (00.007203) mnt: [./sys/kernel/debug](22->21)
> (00.007204) mnt: <--
> (00.007205) mnt: [./sys/fs/cgroup](24->21)
> (00.007205) mnt: <--
> (00.007206) mnt: <--
> (00.007207) mnt: [./acct](23->13)
> (00.007208) mnt: <--
> (00.007208) mnt: [./mnt/asec](25->13)
> (00.007209) mnt: <--
> (00.007210) mnt: [./mnt/obb](26->13)
> (00.007210) mnt: <--
> (00.007211) mnt: [./mnt/shell/emulated](30->13)
> (00.007212) mnt: <--
> (00.007213) mnt: <--
> (00.007214) Collecting netns 2/27061
> (00.007456) ========================================
> (00.007459) Dumping task (pid: 2672)
> (00.007460) ========================================
> (00.007460) Obtaining task stat ...
> (00.007477)
> (00.007478) Collecting mappings (pid: 2672)
> (00.007479) ----------------------------------------
> (00.007596) Collected, longest area occupies 405 pages
> (00.007598) 0x400000-0x595000 (1620K) prot 0x5 flags 0x2 st 0x41 off 0 reg fp shmid: 0
> (00.007600) 0x794000-0x79c000 (32K) prot 0x3 flags 0x2 st 0x41 off 0x194000 reg fp shmid: 0
> (00.007601) 0x79c000-0x7b2000 (88K) prot 0x3 flags 0x22 st 0x201 off 0 reg ap shmid: 0
> (00.007603) 0x85c000-0x87f000 (140K) prot 0x3 flags 0x22 st 0x221 off 0 reg heap ap shmid: 0
> (00.007604) 0x7fe73b638000-0x7fe73b63a000 (8K) prot 0x3 flags 0x21 st 0x101 off 0 reg as shmid: 0x92b13
> (00.007605) 0x7fe73b63a000-0x7fe73b63c000 (8K) prot 0x3 flags 0x21 st 0x101 off 0 reg as shmid: 0x92b10
> (00.007606) 0x7fe73b63c000-0x7fe73b64c000 (64K) prot 0x3 flags 0x22 st 0x201 off 0 reg ap shmid: 0
> (00.007607) 0x7ffc0d131000-0x7ffc0d153000 (136K) prot 0x3 flags 0x122 st 0x201 off 0 reg ap shmid: 0
> (00.007608) 0x7ffc0d16e000-0x7ffc0d170000 (8K) prot 0x1 flags 0x22 st 0x1201 off 0 reg vvar ap shmid: 0
> (00.007609) 0x7ffc0d170000-0x7ffc0d172000 (8K) prot 0x5 flags 0x22 st 0x209 off 0 reg vdso ap shmid: 0
> (00.007610) 0xffffffffff600000-0xffffffffff601000 (4K) prot 0x5 flags 0x22 st 0x204 off 0 vsys ap shmid: 0
> (00.007612) ----------------------------------------
> (00.007614)
> (00.007615) Collecting fds (pid: 2672)
> (00.007615) ----------------------------------------
> (00.007634) Found 11 file descriptors
> (00.007635) ----------------------------------------
> (00.007687) Set up parasite blob using memfd
> (00.007690) Putting parasite blob into 0x7fde89f7a000->0x7fe73b62b000
> (00.007702) Dumping GP/FPU registers for 2672
> (00.007706) Putting tsock into pid 2672
> (00.007734) Wait for parasite being daemonized...
> (00.007737) Parasite 2672 has been switched to daemon mode
> (00.007753) sid=2688 pgid=2672 pid=2672
> (00.007760)
> (00.007760) Dumping opened files (pid: 2672)
> (00.007760) ----------------------------------------
> pie: Error (pie/util-fd.c:92): fcntl(3, F_GETOWN_EX) -> -9
> pie: Error (pie/parasite.c:275): send_fds failed (-1)
> pie: Error (pie/parasite.c:646): Close the control socket for writing
>>
> (00.007768) Error (parasite-syscall.c:829): Can't retrieve FDs from socket
> (00.007770) Error (parasite-syscall.c:305): Message reply from daemon is trimmed (12/0)
> (00.007772) Error (cr-dump.c:1292): Dump files (pid: 2672) failed with -1
> (00.007859) Unlock network
> (00.007863) Unfreezing tasks into 1
> (00.007873) Error (cr-dump.c:1578): Dumping FAILED.
>
> Note that I am using android
>
>
> Best Regard -- ALLAH Hafiz, May Almighty ALLAH bless you
>
>
> --
>
> Abdullah Yousafzai
> https://about.me/yousafzaiabdullah
>
> <https://about.me/yousafzaiabdullah?promo=email_sig>
>
>
> On Tue, Feb 23, 2016 at 11:27 AM, Marcos Lilljedahl <marcosnils at gmail.com <mailto:marcosnils at gmail.com>> wrote:
>
> Abdull,
>
> It's totally possible. Can you please provide the commands and script you're using to run your test?
>
> Marcos.
>
> On Mon, Feb 22, 2016 at 11:09 PM, Abdullah Yousafzai <yousafzai.abdullah at gmail.com <mailto:yousafzai.abdullah at gmail.com>> wrote:
>
> Hello Guys,
>
> I have another question is it possible to re-checkpoint an application i.e.
>
> 1) checkpoint app -> (2) restore it -> (3) than checkpoint again -> (4) than restore again -> (5) than checkpoint again -> ......
>
> I am trying a pattern like this; but when I restored the application at step 2 than I cannot find the app to checkpoint again in ps but I find the criu running (which I am unable to kill/stop).
>
> Note I am using the simplest example (bash script) stated in the CRIU How Tos.
>
> Please comment on the possibilities and fixes if possible.
>
> Thank you all guys.
>
>
> Best Regard -- ALLAH Hafiz, May Almighty ALLAH bless you
>
>
> --
>
> Abdullah Yousafzai
> https://about.me/yousafzaiabdullah
>
> <https://about.me/yousafzaiabdullah?promo=email_sig>
>
>
> _______________________________________________
> CRIU mailing list
> CRIU at openvz.org <mailto:CRIU at openvz.org>
> https://lists.openvz.org/mailman/listinfo/criu
>
>
>
>
>
> _______________________________________________
> CRIU mailing list
> CRIU at openvz.org
> https://lists.openvz.org/mailman/listinfo/criu
>
More information about the CRIU
mailing list