topic Re: Problems after upgrade from E0506 to E0506P03 in IMC

Problems after upgrade from E0506 to E0506P03

hwuensch — Wed, 11 Oct 2017 09:57:42 GMT

Hi,

After upgrading to version E0506P03 some features of the web-interface are broken:

- in "classic" view-mode, after clicking on the "Ressource" tab, the left-sided menu structure is incomplete andcontent is not displayed correctly (empty spaces), see picture 1.

- displaying the "IP View" Ressource triggers an error (s. picture 2). This happens both on classic-mode and desktop-mode.

- oh, and one hint: Do _not_ upgrade to P03 if you have a mysql 5.5 backend: The imcnetresdm process will execute a very complex SQL-Statement on the config_db database after starting up which will take forever on mysql 5.5 (and will prevent the startup of imcnetresdm). After Upgrading to mysql 5.6 this query will be executed in <1sek so everything is fine..

best regards,

hwuensch

Re: Problems after upgrade from E0506 to E0506P03

streeb — Wed, 11 Oct 2017 10:49:16 GMT

hwuensch,

We just upgraded our production instance this week as well and I do not see either of these issues. I am sure you have already but I would open up a support ticket on this one. One thing I do see is that if you do a general search for a device, the result in the dynamic side bar is no longer clickable. This seems to be cross browser issue- I am not sure if the iMC team look at this forum but hopefully they do and can take a look.

Regards

Mike

Re: Problems after upgrade from E0506 to E0506P03

hwuensch — Wed, 11 Oct 2017 12:43:47 GMT

indeed, the "general search" is also failing (it works only one time for me, after that you have to logout/login from the webinterface).

--> we will open a ticket on this issue.

Is anybody else experiencing these frontend errors on E0506P03? We never had such problems bevore, and this time it

was only a minor patch-level update :-(

Regards,

hwuensch

Re: Problems after upgrade from E0506 to E0506P03

NeilR — Sat, 14 Oct 2017 00:29:34 GMT

@hwuensch wrote:
- oh, and one hint: Do _not_ upgrade to P03 if you have a mysql 5.5 backend: The imcnetresdm process will execute a very complex SQL-Statement on the config_db database after starting up which will take forever on mysql 5.5 (and will prevent the startup of imcnetresdm). After Upgrading to mysql 5.6 this query will be executed in <1sek so everything is fine..

Sounds like that's my problem - i'm running MySql 55 and after the update it never finished starting 3 processes - I gave up after waiting 2 hrs and rolled it back - thx

Re: Problems after upgrade from E0506 to E0506P03

NeilR — Tue, 14 Nov 2017 01:54:19 GMT

What was your process for MySql update? on Win or linux?

If I update over the existing ImC install, dbman won't backup or restore and the update process also fails. It runs ok, but otherwise stuck.

Re: Problems after upgrade from E0506 to E0506P03

NeilR — Tue, 21 Nov 2017 23:10:51 GMT

Well the update leaves a query hanging trying to update: Error Code: 1146. Table 'aclm_db.tbl_link' doesn't exist

The query sits there waiting so a couple of processes never start. Killing the query will allow the process to start but a reboot brings it back, and it won't restore the DB anyway.

No P03 until I talk to HP...

Re: Problems after upgrade from E0506 to E0506P03

GLTUKW — Tue, 12 Dec 2017 10:02:55 GMT

I have the same problem after update to 7.3_E0506P03, no ip views .

I use Microsoft SQL server 2008 R2

Re: Problems after upgrade from E0506 to E0506P03

hwuensch — Tue, 12 Dec 2017 10:22:07 GMT

After opening a ticket for this issue:

In our case there was the problem of invalid network masks in 'tbl_subnet' and 'tbl_dev table' (both in databse config_db).

For example, our database contained network mask values of 0.0.80.242 (which is not a valid netmask of course).

You can check that by executing a sql query on the config_db database like this one:

select * from tbl_dev where dev_ip_mask like '0.%';

If this query lists any devices, then you should delete these devices and rediscover them.

(You could also manipulate the wrong entries direct via sql, but beware of side-effects)

Also check tbl_subnet for wrong entries:

select * from tbl_subnet where subnet_mask like '0.%';

there is a change that these or another bad entries will eventually reappear after deleting them: This issue seems to be related with uncomplete/corrupted SNMP-queries to these devices...

Hope this will help you...

Regards,

Helmut

Re: Problems after upgrade from E0506 to E0506P03

NeilR — Wed, 13 Dec 2017 23:45:49 GMT

I did have a couple of devices with 0.0.0.0 subnet masks. And I can confirm that one of them was caused by snmp. After just basic discovery mask was correct. Then set snmp read only, it re-discovered the device but set the mask to 0.0.0.0

No subnets with bad masks

Fixing these did not solve my issue - have a query with a comlex join that runs forever to update a table with under 2000 rows. Stopped all the processes and ran the query directly - but killed it after 15 mins. Some other bad data...