sonic-net/sonic-buildimage

Scripts which perform an installable binary image build for SONiC

sonic-buildimage's Issues

Failed to install SONiC base image (sonic-generic.bin) with ONIE in our x86 switch

In the following commit:
cc7f150#diff-028739c9bdaa77f69d654ce551a93cec

In file installer/x86_64/install.sh, the following line is introduced:
source platforms/$onie_platform

Since we don't need extra cmd for our own platform, I didn't add a file under the "platforms" directory.
Thus the installing failed due to absence of the platform file.

I think it is better to check the existence of file before source it.

The switch will be crashed when inputting "reboot" command under host

Hi all,

I found that there is a trace back issue in master (commitment 1491bf9). When user inputting "reboot" command under host, the switch will be crashed. After investigating, it maybe relates to commitment a877603 (Merge swss and syncd into single service)

The following is the error message.

[  207.772520] BUG: unable to handle kernel paging request at ffffffffa03ff0f0
[  207.780338] IP: [<ffffffff811a746f>] filp_close+0x1f/0x70
[  207.786396] PGD 1816067 PUD 1817063 PMD 27346a067 PTE 0
[  207.792260] Oops: 0000 [#1] SMP 
[  207.795872] Modules linked in: eeprom_mb(O) eeprom w83795 jc42 coretemp bridge stp llc i2c_mux_pca954x i2c_mux i2c_dev i2c_ismt i2c_i801 kvm_intel kvm crc32_pclmul xt_conntrack iTCO_wdt iTCO_vendor_support iptable_filter ipt_MASQUERADE aesni_intel aes_x86_64 xt_addrtype lrw gf128mul glue_helper ablk_helper lpc_ich mfd_core evdev cryptd serio_raw pcspkr iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 ipmi_msghandler tpm_tis tpm button nf_nat_ipv4 shpchp i2c_core nf_nat acpi_cpufreq nf_conntrack processor thermal_sys ip_tables x_tables autofs4 loop ext4 crc16 mbcache jbd2 nls_utf8 nls_cp437 vfat fat aufs(C) squashfs sg sd_mod crc_t10dif crct10dif_generic crct10dif_pclmul crct10dif_common crc32c_intel ahci libahci libata ehci_pci ehci_hcd scsi_mod igb(O) usbcore usb_common dca ptp pps_core [last unloaded: linux_kernel_bde]
[  207.877421] CPU: 1 PID: 1555 Comm: syncd Tainted: G         C O  3.16.0-4-amd64 #1 Debian 3.16.36-1+deb8u2
[  207.896132] task: ffff880036dbe190 ti: ffff8802731bc000 task.ti: ffff8802731bc000
[  207.904518] RIP: 0010:[<ffffffff811a746f>]  [<ffffffff811a746f>] filp_close+0x1f/0x70
[  207.913283] RSP: 0018:ffff8802731bfce0  EFLAGS: 00010246
[  207.919232] RAX: ffffffffa03ff080 RBX: ffff88027349ff00 RCX: 0000000000000027
[  207.927227] RDX: ffff880272466858 RSI: ffff880272466800 RDI: ffff88027349ff00
[  207.935215] RBP: ffff880272466800 R08: ffff8802731bc000 R09: 000000000000b8fe
[  207.943211] R10: 0000000000000001 R11: 0000000000000000 R12: 0000000000000000
[  207.951207] R13: ffff880272466800 R14: 0000000000000001 R15: ffff880272466810
[  207.959205] FS:  00007f4388398740(0000) GS:ffff88027fc80000(0000) knlGS:0000000000000000
[  207.968274] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[  207.974711] CR2: ffffffffa03ff0f0 CR3: 000000027317b000 CR4: 00000000001007e0
[  207.982708] Stack:
[  207.984956]  00000000000002ff 0000000000000027 0000000000000000 ffffffff811c5898
[  207.993262]  ffff880036dbe810 ffff880272a67400 ffff880272a67460 0000000000000005
[  208.001566]  ffff880036fe8760 ffff880036dbe190 ffffffff81069c0e ffff8802731bff58
[  208.009872] Call Trace:
[  208.012602]  [<ffffffff811c5898>] ? put_files_struct+0x78/0xc0
[  208.019142]  [<ffffffff81069c0e>] ? do_exit+0x28e/0xa70
[  208.024997]  [<ffffffff8106a469>] ? do_group_exit+0x39/0xa0
[  208.031241]  [<ffffffff81078928>] ? get_signal_to_deliver+0x1c8/0x5d0
[  208.038461]  [<ffffffff81013492>] ? do_signal+0x42/0xa10
[  208.044412]  [<ffffffff810779c4>] ? do_send_sig_info+0x54/0x70
[  208.050949]  [<ffffffff81013ed8>] ? do_notify_resume+0x78/0xa0
[  208.057488]  [<ffffffff81518a8a>] ? int_signal+0x12/0x17
[  208.063437] Code: 66 66 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 41 54 55 53 48 8b 47 38 48 89 fb 48 85 c0 74 44 48 8b 47 28 45 31 e4 48 89 f5 <48> 8b 40 70 48 85 c0 74 05 ff d0 41 89 c4 f6 43 45 40 75 16 48 
[  208.084896] RIP  [<ffffffff811a746f>] filp_close+0x1f/0x70
[  208.091039]  RSP <ffff8802731bfce0>
[  208.094942] CR2: ffffffffa03ff0f0
[  208.098652] ---[ end trace 7dcc5bf3c94d2437 ]---

Does anyone encounter this issue?

README.md leaves out docker-database

I caught a question on "what is docker-daabase, is this different from docker-base?"

docker-database is not in the list of images in the readme.

docker-base not squashed

missing squash

## Flatten the image by importing an exported container on this image
## Note: it will squash the image with only one layer and lost all metadata such as ENTRYPOINT,
##       so apply only to the base image
## TODO: wait docker-squash supporting Docker 1.10+
## ref: https://github.com/jwilder/docker-squash/issues/45
if [ "$docker_image_name" = "docker-base" ]; then
    ## Run old image in a container
    tmp_container=$(docker run -d ${docker_image_name} /bin/bash)
    ## Export the container's filesystem, then import as a new image
    docker export $tmp_container | docker import - ${docker_image_name}
    ## Remove the container
    docker rm -f $tmp_container || true
    ## Remove the old image
    docker rmi -f $image_id || true
fi

teamd: Error: either "dev" is duplicate, or "boot" is a garbage.

This is because the folder /etc/teamd is empty.
Need to check if /etc/teamd folder is empty or not.
https://github.com/Azure/sonic-buildimage/pull/441/files could resolve this issue.
But line https://github.com/marian-pritsak/sonic-buildimage/blob/97a59f15d5c57b83e604efd50cb674a818f6fec9/dockers/docker-teamd/start.sh#L18 will still return /etc/teamd/* when the folder is empty.

pack saiserver into a .deb archive in Makefile

[swss] 'oper_status' attribute can have a value of '2'?

The 'oper_status' attribute of a port in our DB has three potential values: 'up', 'down' and '2'.

Can reproduce on a T0 setup. Simply 'ifdown' a port that's part of the VLAN. Reading the database rapidly, you will see the 'oper_status' change from 'up' to 'down' and then quickly to '2', where it will stay.

Likewise, bringing the port back up using 'ifup', you will see the 'oper_status' change from '2' to 'down' and then to 'up'.

configurations are re-generated across reboots

@taoyl-ms Sometimes people would like to update some configurations on the device by themselves. but rebooting will trigger the files inside the dockers to be re-generated again via the minigraph.xml. Is it by design or shall we able to leave the configurations there before manually regenerating this configurations?

Remove docker-bgp?

Is docker-bgp deprecated? Should/can it be removed from the project?

redirect commands in /etc/network/interfaces to /dev/null

commands such as brctl and teamdctl
also append || true at the end of the commands

when flapping the link using ifup/ifdown, command may fail and the interface cannot be brought up.

No such file or directory #include <switch_sai_rpc_server.h>

syncd.h:20:35: fatal error: switch_sai_rpc_server.h: No such file or directory
 #include <switch_sai_rpc_server.h>

Build failure on p4 platform:
https://sonic-jenkins.westus.cloudapp.azure.com/job/p4/job/buildimage-p4-all-pr/649/console

The order to apply qos.json and buffers.json cannot be reversed

Currently there are internal dependencies in between qos.json and buffers.json files that if the order of applying the configurations cannot be guaranteed, some configurations cannot be applied correctly.

After brief debugging on this issue, it is possible that there might be some internal dependencies in bcmsai package.

thrift pacakge test failure

https://sonic-jenkins.westus.cloudapp.azure.com/job/broadcom/job/buildimage-brcm-all/45/console

redis 3.2.4 build fails intermitantly

https://sonic-jenkins.westus.cloudapp.azure.com/job/cavium/job/buildimage-cavm-all-pr/12/console

add an option to specify the baud rate for sonic

add an option to specify the baud rate for sonic, or check if it is possible to use the same baudrate as the onie image in long run.

Build will fail when working on a branch name with "/"

git_revision will contain "/", and later sed will fail.

Example log:

Build ONIE installer
+ arch=x86_64
+ machine=broadcom
+ platform=x86_64-broadcom-r0
+ installer_dir=installer
+ platform_conf=platform/broadcom/platform.conf
+ output_file=target/sonic-broadcom.bin
+ demo_type=OS
+ git_revision=taoyl/test.0-c8a315e
+ onie_image_part_size=32768
+ shift 9
+ [ ! -d installer ]
+ [ ! -r installer/sharch_body.sh ]
+ [ ! -d installer/x86_64 ]
+ [ ! -r installer/x86_64/install.sh ]
+ [ -n taoyl/test.0-c8a315e ]
+ [ -n 32768 ]
+ [ -r platform/broadcom/platform.conf ]
+ [ 1 -gt 0 ]
+ tmp_dir=
+ echo -n Building self-extracting install image .
Building self-extracting install image .+ mktemp --directory
+ tmp_dir=/tmp/tmp.85lbt3lbEE
+ tmp_installdir=/tmp/tmp.85lbt3lbEE/installer
+ mkdir /tmp/tmp.85lbt3lbEE/installer
+ cp -r installer/x86_64/install.sh installer/x86_64/macset.sh installer/x86_64/platforms /tmp/tmp.
85lbt3lbEE/installer
+ cp onie-image.conf /tmp/tmp.85lbt3lbEE/installer
+ echo
+ sed -e s/[\/&]/\\&/g
+ EXTRA_CMDLINE_LINUX=
+ sed -i -e s/%%DEMO_TYPE%%/OS/g -e s/%%GIT_REVISION%%/taoyl/test.0-c8a315e/g -e s/%%ONIE_IMAGE_PART
_SIZE%%/32768/ -e s/%%EXTRA_CMDLINE_LINUX%%// /tmp/tmp.85lbt3lbEE/installer/install.sh
sed: -e expression #2, char 26: unknown option to `s'
+ clean_up 1
+ rm -rf /tmp/tmp.85lbt3lbEE
+ exit 1
slave.mk:314: recipe for target 'target/sonic-broadcom.bin' failed
make: Leaving directory '/sonic'
make: *** [target/sonic-broadcom.bin] Error 1
Makefile:33: recipe for target 'target/sonic-broadcom.bin' failed
make: *** [target/sonic-broadcom.bin] Error 2

[config engine] Give more detail in error message upon illegal minigraph

[jenkins] make broadcom all failure

cannot build onie and aboot image for broadcom at the same time.

https://sonic-jenkins.westus.cloudapp.azure.com/job/broadcom/job/buildimage-brcm-all/74/console

rc.local fails on boot

● rc-local.service - /etc/rc.local Compatibility
Loaded: loaded (/lib/systemd/system/rc-local.service; static)
Active: failed (Result: exit-code) since Thu 2017-03-02 12:01:05 UTC; 4min 35s ago
Process: 439 ExecStart=/etc/rc.local start (code=exited, status=1/FAILURE)

Mar 02 12:01:05 arc-switch1025 rc.local[439]: install platform dependent packages at the first boot time
Mar 02 12:01:05 arc-switch1025 rc.local[439]: cp: cannot stat '/usr/share/sonic/device//minigraph.xml': No such file or directory
Mar 02 12:01:05 arc-switch1025 systemd[1]: rc-local.service: control process exited, code=exited status=1
Mar 02 12:01:05 arc-switch1025 systemd[1]: Failed to start /etc/rc.local Compatibility.
Mar 02 12:01:05 arc-switch1025 systemd[1]: Unit rc-local.service entered failed state.

Remove platform/ subdirectory from sonic-config-engine

1.) Confirm all components are now referencing device-specific data in the device/ directory, and that the platform/ subdirectory in sonic-config-engine is truly deprecated.

2.) Remove platform/ subdirectory from sonic-config-engine.

Impossible to build platform independent docker container

From the documentation:

Usage

To build NOS installer image and docker images, run command line

make configure PLATFORM=[VENDOR]

What if we build a docker container which doesn't depend on the platform?

omem=1430033368

admin@str-s6000-acs-7:~$ redis-cli client list | grep -v omem=0 
id=96 addr=/var/run/redis/redis.sock:0 fd=40 name= age=162949 idle=162940 flags=U db=0 sub=0 psub=1 multi=-1 qbuf=0 qbuf-free=0 obl=16351 oll=83587 omem=1430033368 events=rw cmd=psubscribe
admin@str-s6000-acs-7:~$ show version

SONiC Software Version: SONiC.master.0-c7d540c
Distribution: Debian 8.7
Kernel: 3.16.0-4-amd64
Build commit: c7d540c
Build date: Sat Apr 29 09:56:25 UTC 2017
Built by: lgh@lgh-issaquah

Docker images:
REPOSITORY                TAG                 IMAGE ID            SIZE
docker-syncd-brcm         latest              e91056044328        282.9 MB
docker-orchagent-brcm     latest              5d6e3f1eeec2        252.8 MB
docker-lldp-sv2           latest              6362b759dd9c        254.9 MB
docker-dhcp-relay         latest              662c2975b024        248.5 MB
docker-database           latest              544a69626fd3        216.4 MB
docker-snmp-sv2           latest              301f15e9b299        290 MB
docker-teamd              latest              8dbd472b5815        250.3 MB
docker-platform-monitor   latest              29412eed58bf        273.5 MB
docker-fpm-quagga         latest              63517bd3a437        256.9 MB

admin@str-s6000-acs-7:~$ docker stop lldp
lldp

admin@str-s6000-acs-7:~$ redis-cli client list | grep -v omem=0

dockers/docker-snmp-sv2/config.sh still references deprecated alias_map.json file

Line 12 of dockers/docker-snmp-sv2/config.sh is trying to copy deprecated alias_map.json file:

/bin/cp -rf /usr/share/sonic/$hwsku/alias_map.json /etc/snmp/

config gen mismatch for docker-snmp-sv2

snmpd is started by supervisord, not sure if /etc/default/snmp is still needed.

bgp container not started due to database

admin@str-s6000-acs-7:~$ sudo systemctl status bgp
â—� bgp.service - BGP container
   Loaded: loaded (/etc/systemd/system/bgp.service; enabled)
   Active: failed (Result: exit-code) since Wed 2017-03-08 10:55:45 UTC; 21h ago
 Main PID: 802 (code=exited, status=125)

Mar 08 10:55:39 str-s6000-acs-7 systemd[1]: Started BGP container.
Mar 08 10:55:42 str-s6000-acs-7 bgp.sh[802]: docker: Error response from daemon: No such container: database.
Mar 08 10:55:42 str-s6000-acs-7 bgp.sh[802]: See 'docker run --help'.
Mar 08 10:55:42 str-s6000-acs-7 systemd[1]: bgp.service: main process exited, code=exited, status=125/n/a
Mar 08 10:55:45 str-s6000-acs-7 bgp.sh[903]: Error response from daemon: No such container: bgp
Mar 08 10:55:45 str-s6000-acs-7 systemd[1]: bgp.service: control process exited, code=exited status=1
Mar 08 10:55:45 str-s6000-acs-7 systemd[1]: Unit bgp.service entered failed state.

[sonic-cfggen:] refactor and move parse_port_config to sonic_platform.py

A default dhcp_relay.yml is missing from oneimage

Will cause dhcp_relay docker failing to start.
@jleveque

disklabel can only be 16 char in ext4, now it is too long

Creating new SONiC-OS partition /dev/sda3 ...
Could not create partition 3 from 788480 to 67897343
Unable to set partition 3's name to 'SONiC-OS-HEAD.751-701c1eb'!
Error encountered; not saving changes.
Warning: The first trial of creating partition failed, trying the largest aligned available block of sectors on the disk
Warning: The kernel is still using the old partition table.
The new table will be used at the next reboot.
The operation has completed successfully.
mke2fs 1.42.8 (20-Jun-2013)
Discarding device blocks: done                            
Filesystem label=SONiC-OS-HEAD.75
OS type: Linux

Filesystem label=SONiC-OS-HEAD.75

This causes issue in booting.

Booting `SONiC-OS-HEAD.751-701c1eb'

error: no such device: SONiC-OS-HEAD.751-701c1eb.
Loading SONiC-OS-HEAD.751-701c1eb OS kernel ...
Loading SONiC-OS-HEAD.751-701c1eb OS initial ramdisk ...

latest sx-kernel package is missing

@marian-pritsak

I cannot find the sx-kernel image, we need to pull it and upload it to the sonic apt-get repo.

sx-kernel_1.mlnx.4.2.3002_amd64.deb

Crash on changing of IP address on a network interface with an active neighbor

Topology
TG-DUT-DUT-TG

Steps

1. Boot two switches with SONIC (OCP branch) and minigraph
2. sudo ifconfig Ethernet0 2.2.2.1 netmask 255.255.255.0 on any switch

Observed results
Feb 10 14:23:23 OCPSCH0104001MS NOTICE orchagent: :- removeSubnetRoute: Remove subnet route to 10.10.1.2/30 from Ethernet0
Feb 10 14:23:23 OCPSCH0104001MS NOTICE orchagent: :- removeIp2MeRoute: Remove packet action trap route ip:10.10.1.2
Feb 10 14:23:23 OCPSCH0104001MS NOTICE orchagent: :- removeRouterIntfs: Remove router interface for port Ethernet0
Feb 10 14:23:23 OCPSCH0104001MS ERR orchagent: :- meta_sai_validate_oid: oid is set to null object id
Feb 10 14:23:23 OCPSCH0104001MS ERR orchagent: :- removeRouterIntfs: Failed to remove router interface for port Ethernet0, rv:-5
Feb 10 14:23:23 OCPSCH0104001MS ERR orchagent: :- main: Failed due to exception: Failed to remove router interface.
Feb 10 14:23:23 OCPSCH0104001MS INFO docker[1068]: terminate called without an active exception
Feb 10 14:23:23 OCPSCH0104001MS INFO kernel: [ 1264.781525] Core dump to |/usr/bin/coredump-compress orchagent 91 pipe failed

[dockers]: "docker stop" command fails on dockers with shell entrypoints

docker stop sends a signal to the process with PID of 1 inside the docker. However, if the docker has an entrypoint of /bin/bash running one or more commands, /bin/bash becomes PID 1 and the signal will be received by /bin/bash but it will not be forwarded along to processes running in the shell. This causes docker stop to timeout instead of shutting the docker down properly.

Stop jobs take extremely long on shutdown/reboot

Upon reboot (shutdown) a few stop jobs take a very long time to complete. The following are known culprits:

• Switch state service
• Docker Application Container Engine

portstat fails

# portstat 
Traceback (most recent call last):
  File "/usr/bin/portstat", line 320, in <module>
    main()
  File "/usr/bin/portstat", line 273, in main
    cnstat_dict = portstat.get_cnstat()
  File "/usr/bin/portstat", line 77, in get_cnstat
    cnstat_dict[port] = get_counters(counter_port_name_map[port])
  File "/usr/bin/portstat", line 67, in get_counters
    fields[pos] += int(self.db.get(self.db.COUNTERS_DB, full_table_id, counter_name))
TypeError: int() argument must be a string or a number, not 'NoneType'

Reproduced on Mellanox platform.
Can anyone try it on another to help locate the problem?

Failure trying to run: chroot /sonic-buildimage/fsroot mount -t proc proc /proc

Hello,
I'm trying to build an image for Broadcom.

Step 1-3 is fine, the problem appears when I'm using make command in usage section.
Docker version 1.12.1, build 6f9534c

Error output:

I: Extracting util-linux...
I: Extracting liblzma5...
I: Extracting zlib1g...
W: Failure trying to run: chroot /sonic-buildimage/fsroot mount -t proc proc /proc
W: See /sonic-buildimage/fsroot/debootstrap/debootstrap.log for details
Makefile:141: recipe for target 'target/sonic-generic.bin' failed
make: *** [target/sonic-generic.bin] Error 1
rm src/initramfs-tools/initramfs-tools_0.120_all.deb src/sonic-linux-kernel/linux-image-3.16.0-4->amd64_3.16.7-ckt11-2+acs8u2_amd64.deb

Thanks.

teamsyncd fails on start

● teamd.service - TEAMD container
Loaded: loaded (/etc/systemd/system/teamd.service; enabled)
Active: active (running) since Fri 2017-03-03 15:05:23 UTC; 4min 51s ago
Process: 14282 ExecStop=/usr/bin/teamd.sh stop (code=exited, status=0/SUCCESS)
Main PID: 14315 (teamd.sh)
CGroup: /system.slice/teamd.service
├─14315 /bin/bash /usr/bin/teamd.sh start
└─14324 docker start -a teamd

Mar 03 15:05:26 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:26 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:26 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:26 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:26 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:27 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:27 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:27 arc-switch1026 teamd.sh[14315]: This program is not intended to be run as root.
Mar 03 15:05:27 arc-switch1026 teamd.sh[14315]: terminate called after throwing an instance of 'std::system_error'
Mar 03 15:05:27 arc-switch1026 teamd.sh[14315]: what(): Unable to connect to redis (unixs-socket): Cannot assign requested address

systemd-journald[224]: [/etc/systemd/journald.conf:21] Failed to parse size value, ignoring:

fsck from util-linux 2.25.2
/dev/sda1: clean, 21053/122400 files, 193789/489148 blocks
[ 16.049420] systemd-journald[224]: [/etc/systemd/journald.conf:18] Failed to parse size value, ignoring: 50MB
[ 16.167920] systemd-journald[224]: [/etc/systemd/journald.conf:21] Failed to parse size value, ignoring: 50MB
[ OK ] Started LSB: Raise network interfaces..

umount: /proc: target is busy

https://sonic-jenkins.westus.cloudapp.azure.com/job/common/job/buildimage-baseimage-pr/385/console

[INFO] Umount all
+ sudo LANG=C chroot ./fsroot fuser -km /proc
/proc:                6095
+ sudo LANG=C chroot ./fsroot umount /proc
umount: /proc: target is busy
        (In some cases useful info about processes that
         use the device is found by lsof(8) or fuser(1).)
+ rm -f /tmp/tmp.AjVwDYDeYQ
+ sudo umount ./fsroot/proc
umount: /sonic/fsroot/proc: target is busy
        (In some cases useful info about processes that
         use the device is found by lsof(8) or fuser(1).)
+ true
+ true
slave.mk:303: recipe for target 'target/sonic-generic.bin' failed
make: Leaving directory '/sonic'
make: *** [target/sonic-generic.bin] Error 1
Makefile:32: recipe for target 'target/sonic-generic.bin' failed
make: *** [target/sonic-generic.bin] Error 2
Build step 'Execute shell' marked build as failure
Adding one-line test results to commit status...
Setting status of ba493b72e4a0fb0170fc0f4c51488641e8da6658 to FAILURE with url https://sonic-jenkins.westus.cloudapp.azure.com/job/common/job/buildimage-baseimage-pr/385/ and message: ' No test results found.'
Using context: baseimage
Finished: FAILURE

P4 bulilds failure undeterministically

The test part fails randomly.

test_suite
  ThriftTestCase
    testByte ...                                                           [OK]
    testDouble ...                                                         [OK]
    testException ...                                                      [OK]
    testI32 ...                                                            [OK]
    testI64 ...                                                            [OK]
    testOneway ...                                                       [FAIL]
    testString ...                                                         [OK]
    testStruct ...                                                         [OK]
    testVoid ...                                                           [OK]

generic all build failure

https://sonic-jenkins.westus.cloudapp.azure.com/job/common/job/buildimage-generic-all/

integrate docker-ptf-saithrift, docker-saiserver into buildimage

we need vendor specific dockers for docker-ptf-saithrift and docker-saiserver.

minigraph.py crashed when no png is in the minigraph

When there's no png info, minigraph.py will crash, because console_dev and mgmt_dev variables are missing. Better to have some basic checks here.

Besides, if the role tag is missing, the bgpd.conf.j2 will fail to be deployed.

Remove docker-basic_router?

Is docker-basic_router deprecated? Should/can it be removed from the project?

Remove unique_name filter in sonic-cfggen and update interfaces.j2

With the pull request #408, we could remove some changes introduced in #381 so that to make the code easier to read.

change SED_DPKG to a jinji2 template

https://github.com/kolypto/j2cli for templating

docker connect failure in the build

https://sonic-jenkins.westus.cloudapp.azure.com/job/mellanox/job/buildimage-mlnx-all-pr/277/console

• BUILD_TEMPLATES=files/build_templates
• IMAGE_CONFIGS=files/image_config
• trap_push clean_sys
• local oldcmd=true
• local 'newcmd=clean_sys; true'
• trap -- 'clean_sys; true' EXIT INT TERM HUP
• _trap_push 'clean_sys; true'
• local 'next=clean_sys; true'
  ++ sed -e 's/'''/'''\''''''/g'
  ++ echo 'clean_sys; true'
• eval 'trap_push() {
  local oldcmd='''clean_sys; true'''
  local newcmd="$1; $oldcmd"
  trap -- "$newcmd" EXIT INT TERM HUP
  _trap_push "$newcmd"
  }'
• sudo LANG=C chroot ./fsroot mount sysfs /sys -t sysfs
• sudo chroot ./fsroot service docker start
  mount: cgroup is already mounted or /sys/fs/cgroup/cpu busy
  cgroup is already mounted on /sys/fs/cgroup
  mount: cgroup is already mounted or /sys/fs/cgroup/cpuacct busy
  cgroup is already mounted on /sys/fs/cgroup
  mount: cgroup is already mounted or /sys/fs/cgroup/net_cls busy
  cgroup is already mounted on /sys/fs/cgroup
  mount: cgroup is already mounted or /sys/fs/cgroup/net_prio busy
  cgroup is already mounted on /sys/fs/cgroup
  Starting Docker: docker.
• sudo chroot ./fsroot docker version
  Client:
  Version: 1.11.1
  API version: 1.23
  Go version: go1.5.4
  Git commit: 5604cbe
  Built: Tue Apr 26 23:11:07 2016
  OS/Arch: linux/amd64
  Cannot connect to the Docker daemon. Is the docker daemon running on this host?

LAGs are not updated correctly after teamd restart

Reproduced on:

build_version: HEAD.211-5585221
debian_version: 8.7
kernel_version: 3.16.0-4-amd64
asic_type: mellanox
build_date: Wed Apr 12 05:07:40 UTC 2017
built_by: johnar@jenkins-worker-1

Steps to reproduce:

$ systemctl restart teamd

After I restart teamd, some LAGs are not working anymore. From SDK dump I can see that part of them have no members on hardware, some have only one, and some have two.

Including generic platform by default breaks p4

p4 platform defines a single docker with all packages.
However, after introducing generic platform, docker-lldp, docker-database etc. are also built.

dpkg-query: error: failed to open package info file `/var/lib/dpkg/updates/0001' for reading: No such file or directory

dh_perl -plibnl-3-200-udeb
dh_shlibdeps -plibnl-3-200-udeb
dpkg-query: error: failed to open package info file `/var/lib/dpkg/updates/0001' for reading: No such file or directory
dpkg-shlibdeps: error: dpkg-query --control-path libc6:amd64 shlibs gave error exit status 2
dh_shlibdeps: dpkg-shlibdeps -Tdebian/libnl-3-200-udeb.substvars -tudeb debian/libnl-3-200-udeb/lib/libnl-3.so.200.22.0 returned exit code 2
/usr/share/cdbs/1/rules/debhelper.mk:277: recipe for target 'binary-predeb-IMPL/libnl-3-200-udeb' failed
make[2]: *** [binary-predeb-IMPL/libnl-3-200-udeb] Error 2
make[2]: Leaving directory '/var/build/workspace/mellanox/buildimage-mlnx-all-pr/src/libnl3/libnl3'

docker-fpm and docker-team depend on libsai

docker-fpm needs fpmsyncd. docker-team needs teamsyncd.
these two binaries only depends on libswsscommon but they are in swss repository.
swss repository depends on libsairedis which is in sairedis repo.
sairedis repo depends on libsai because syncd is in sairedis repo.

thus, docker-fpm and docker-team depend on libsai.

the ideal scenario is that only fpmsyncd and teamsyncd binaries are inside the docker. no libsairedis/libsai/swss are installed.

Jenkins builds failing due to lack of free space

https://sonic-jenkins.westus.cloudapp.azure.com/job/common/job/buildimage-baseimage-pr/413/:

make[4]: Entering directory '/sonic/src/sonic-linux-kernel/linux-3.16.36'
dh_installdocs 
dh_installchangelogs
dh_strip
dh_compress
dh_fixperms
dh_installdeb
dh_gencontrol -- -Vkernel:Recommends=irqbalance,
dh_md5sums
dh_builddeb -- -Zxz 
dpkg-deb: building package `linux-image-3.16.0-4-amd64' in `../linux-image-3.16.0-4-amd64_3.16.36-1+deb8u2_amd64.deb'.
make[4]: Leaving directory '/sonic/src/sonic-linux-kernel/linux-3.16.36'
dpkg-deb (subprocess): compressing data member: lzma write error: No space left on device
dpkg-deb: error: subprocess <compress> from tar -cf returned error exit status 2
dh_builddeb: dpkg-deb -Zxz --build debian/linux-image-3.16.0-4-amd64-dbg .. returned exit code 2
debian/rules.real:173: recipe for target 'install-base' failed

https://sonic-jenkins.westus.cloudapp.azure.com/job/broadcom/job/buildimage-brcm-all-pr/83/:

running build
running build_ext
error: [Errno 28] No space left on device: 'build/temp.linux-x86_64-3.6-pydebug'
Makefile:627: recipe for target 'sharedmods' failed
make[3]: *** [sharedmods] Error 1

https://sonic-jenkins.westus.cloudapp.azure.com/job/cavium/job/buildimage-cavm-all-pr/410/:

/tmp/ccJe73Wu.s: Assembler messages:
/tmp/ccJe73Wu.s: Fatal error: can't write tests-converter_ut.o: No space left on device