SRX

Hi,
I am planning to do Junos upgarde SRX 240H from 11.4R7.5 to 12.1X44-D35.5 ,

Setup: Chassis cluster - node1 as primary for all redundancy groups, node0 as secondary for all groups.

As juniper advised Simultaneously, at the same time, reboot both

I am woundering on which node should enter the reboot command first, so I will not lock out my session if redudancy group failover.


======
NODE1> show chassis cluster status
Cluster ID: 1
Node                  Priority          Status    Preempt  Manual failover

Redundancy group: 0 , Failover count: 1
    node0                   100         secondary      no       no  
    node1                   1           primary        no       no  

Redundancy group: 1 , Failover count: 3
    node0                   100         secondary      no       no  
    node1                   1           primary        no       no  

{primary:node1}
NODE1>

Please also confirm if the below is the recommended release and if any one is using this witout any issues.

JTAC Recommended release for this product is: 12.1X44-D35.5

http://www.juniper.net/support/downloads/?p=srx240#sw

Hi Mass,

   Do you have console access of these devices? If yes then you can do as below

1) You can install Junos on both devices without reboot ==> request system software add no-copy no-validate <path of package>

2) After installation, halt node 0   => request system halt

3) Once node 0 halted, then you can reboot node 1  ==> request system reboot

3) Once node 1 come back in 12.1X44-D35 version, make sure all good and then you can boot up node 0 by accessing its console and hit enter.

4) Node 0 will boot up with 12.1X44-D35 and join the cluster.

Yes, 12.1X44-D35 should be stable release.

Thanks,

SHKM

Hi Mass,

yes 12.1X44-D35 is the recommended release.

After uploading Junos code to both devices , both nodes needs to be rebooted simultaneously within few second interval.

As per RG group status, have console connection to both devices and reboot node1 first and then Node0.

If you do not have console session , then reboot node0 and then Node1 but Node0 will be primary after reboot.

Whichever node is rebooted first , it will come online first and become primary.

Thanks & Regards,

rparthi

[Please Mark My Solution Accepted if it Helped, Kudos are Appreciated Too] .....

Hello

You could perform in-service upgrade where reboot of both nodes simultaneously is not required.

Importantly, there should not be any outage as well.

Reference:

http://www.juniper.net/techpubs/en_US/junos12.1/topics/task/operational/chassis-cluster-upgrading-and-aborting-backup-and-primary-device-with-icu.html

Regards,

Raveen

Hi All,

Many thanks for giving your kind attention.

All solutions looks work able and I will do same lab test before actual producation upgrade

Thanks again for your assistance

Hi

I am worry about the amount of free disk space necessary to upgrade My device with a new version of the Junos OS

that the /var filesystem is low on free disk space

I understand version 12 is about 143M

/Downloads$ ls -sh junos-srxsme-12.1X44-D35.5-domestic.tgz
143M junos-srxsme-12.1X44-D35.5-domestic.tgz
/Downloads$

I want to avoid deleting files using request system storage cleanup

Any advise ?

=======

node0:
--------------------------------------------------------------------------
Filesystem       1024-blocks       Used      Avail   Capacity   Mounted on
/dev/da0s1a           299052     142280     132848        52%  /
devfs                      1          1          0       100%  /dev
/dev/md0              377210     377210          0       100%  /junos
/cf                   299052     142280     132848        52%  /junos/cf
devfs                      1          1          0       100%  /junos/dev/
procfs                     4          4          0       100%  /proc
/dev/bo0s3e            24876         90      22796         0%  /config
/dev/bo0s3f           350636     147122     175464        46%  /cf/var
/dev/md1              171860      18734     139378        12%  /mfs
/cf/var/jail          350636     147122     175464        46%  /jail/var
/cf/var/log           350636     147122     175464        46%  /jail/var/log
devfs                      1          1          0       100%  /jail/dev
/dev/md2               40172          4      36956         0%  /mfs/var/run/utm
/dev/md3                1884          4       1730         0%  /jail/mfs

node1:
--------------------------------------------------------------------------
Filesystem       1024-blocks       Used      Avail   Capacity   Mounted on
/dev/da0s1a           299052     142096     133032        52%  /
devfs                      1          1          0       100%  /dev
/dev/md0              377210     377210          0       100%  /junos
/cf                   299052     142096     133032        52%  /junos/cf
devfs                      1          1          0       100%  /junos/dev/
procfs                     4          4          0       100%  /proc
/dev/bo0s3e            24876        100      22786         0%  /config
/dev/bo0s3f           350636     218848     103738        68%  /cf/var
/dev/md1              171860      18802     139310        12%  /mfs
/cf/var/jail          350636     218848     103738        68%  /jail/var
/cf/var/log           350636     218848     103738        68%  /jail/var/log
devfs                      1          1          0       100%  /jail/dev
/dev/md2               40172          4      36956         0%  /mfs/var/run/utm
/dev/md3                1884        356       1378        21%  /jail/mfs

{primary:node1}

Hi Mass,

You can follow upgrade using "USB". In this case we mount USB to SRX and there should not be an issue with memory.

Mode details on this method can be found in http://kb.juniper.net/InfoCenter/index?page=content&id=KB16652

Thanks,

Suraj

Please Mark My Solution Accepted if it Helped, Kudos are Appreciated tooj

Hi Suraj,

Thanks for USB solution,

I am not allowed to plug externeal devices, as its customer firewall.

I am woundering if below files are save to delete and will i get enough space for new Junos version 12.1X44-D35.5.

% df -h /var/tmp/
Filesystem     Size    Used   Avail Capacity  Mounted on
/dev/bo0s3f    342M    218M     97M    69%    /cf/var
% cd /var/tmp/
% ls -lrt
total 48
drwxrwxrwt  2 root  wheel   512 Feb 11  2013 vi.recover
drwxrwxrwx  2 root  wheel   512 Feb 11  2013 pics
drwxrwxrwx  2 root  wheel   512 Feb 11  2013 install
-rw-r--r--  1 root  wheel     0 Feb 11  2013 spu_kmd_init
drwxr-xr-x  2 root  wheel   512 Feb 11  2013 gres-tp
drwxr-xr-x  2 root  wheel   512 Feb 11  2013 rtsdb
-rw-r--r--  1 root  wheel    34 Sep  2  2013 gksdchk.log
-rw-r--r--  1 root  wheel  3811 Sep  2  2013 cleanup-pkgs.log
-rw-r--r--  1 root  wheel     4 Sep  2  2013 idp_license_info
-rw-r--r--  1 root  wheel   155 Sep  2  2013 krt_gencfg_filter.txt
-rw-r--r--  1 root  wheel     0 Jun 18 14:22 eedebug_bin_file
-rw-r--r--  1 root  wheel    60 Jun 18 14:23 policy_status
drwxr-xr-x  3 root  wheel   512 Jul  9 23:00 sec-download
%

Thanks

Masroor

Hi Masroor,

Yes, we can delete all the files under /var/tmp/ safely.

You can also take a backup of /var/log/ and delete them, and deleting system software backup as below.

> request system software delete-backup 
Delete backup system software package [yes,no] (no) yes

Thanks,

Suraj

If this worked for you please flag my post as an "Accepted Solution" so others can benefit. A kudo would be cool if you think I earned it.

Hi Suraj,

I think,  will avoid doing this as if i will software back then  will not be able to roll-back.

> request system software delete-backup 
Delete backup system software package [yes,no] (no) yes

Thanks

Mas

Hi Mass,

You can run the following command (du -a /cf |sort -n -r | more) from shell prompt and delete the large file which are not needed.

once you identify the file that you need to delete , then cd to the directory and use rm filename to delete them manually.

Regards
rparthi
 

Kudos are Always Appreciated.....