Re: BMC Atrium Core 8.1.02 Installer
Hi Jason
Yeah I have been looking everywhere to find a possible answer to this
problem. Currently I have logged this with BMC Support as my version 8
upgrade project is now waiting for this to be resolved.
I have tried numerous ways to try and sort this out, from restoring the DB
and re-running the ARS core 8.1.02 Upgrade (which went through with no
errors) to then re-running the Atrium 8.1.02 upgrade.
The error message we are getting when running Atrium Installer SP2 with a
installation type of Upgrade is:
THROWABLE EVENT {Description=[Failed to execute Rule Engine
2],Detail=[[ERROR][Wed Oct 22 11:33:01.383] ImportFileNode- ARImport() for
COM%LoadCompany2.def returned non-zero return code 2
[ERROR][Wed Oct 22 11:33:01.383] ImportFileNode- 8000 The error handler can
not be found on this server.
[WARNING][Wed Oct 22 11:33:01.383] ImportFileNode- 55 The following item
was not imported COM:DCL:CreateCOMCompany_740`!
[ERROR][Wed Oct 22 11:33:01.383] LoadComponent- Definition Import failed
for C:\Program Files\BMC
Software\AtriumCore\dsl\com\workflow\en\.\COM%LoadCompany2.def (return code
2)
[ERROR][Wed Oct 22 11:33:13.215] LoadComponent- At least one file in
definitions list returned an error
[ERROR][Wed Oct 22 11:33:21.471] LoadComponent- Errors processing component
com, worst overall return code = 2]}
I have tried to import that DEF file manually through the developer tool
and I get the same error. So next step was to try and import the DEF file
that contained the Error Handler which imported with no problems. Tried to
import the DEF file again (COM:LoadCompan2.def) and still get the same
error.
On Tue, Oct 21, 2014 at 8:21 PM, Jason Miller jason.mil...@gmail.com
wrote:
**
Have you checked the kb.bmc.com or communities.bmc.com? I remember
seeing a write up on this somewhere. I think it was one of the webinars in
the Atrium Webinar Series https://communities.bmc.com/docs/DOC-22363.
Jason
On Mon, Oct 20, 2014 at 5:02 AM, BradRemedy bradrem...@gmail.com wrote:
**
Hi
So just wanted to share one of the errors in the CMDB installation log:
*[ERROR][Mon Oct 20 13:21:36.930] ImportFileNode- 8000 The error handler
can not be found on this server. *
*[WARNING][Mon Oct 20 13:21:36.930] ImportFileNode- 55 The following item
was not imported COM:DCL:CreateCOMCompany_740`!*
*[ERROR][Mon Oct 20 13:21:36.931] LoadComponent- Definition Import failed
for C:\Program Files\BMC
Software\AtriumCore\dsl\com\workflow\en\.\COM%LoadCompany2.def (return code
2)*
*[ERROR][Mon Oct 20 13:21:48.381] LoadComponent- At least one file in
definitions list returned an error*
*[ERROR][Mon Oct 20 13:21:56.310] LoadComponent- Errors processing
component com, worst overall return code = 2]}*
I went and tried to import that Def file using the Developer tool and got
the same error which is The following item was not imported:
COM:DCL:CreateCOMCompany_740`!'
I will continue to investigate :-)
Cheers
Brad
On Mon, Oct 20, 2014 at 1:26 PM, Patil, Vivek vivek_pa...@bmc.com
wrote:
**
Hi Brad,
Please check what is failed in 8.1 SP1 because of which you see this
error during upgrade.
Please take DB backup .
One will have to fix the failure.
Thanks,
Vivek
*From:* Action Request System discussion list(ARSList) [mailto:
arslist@ARSLIST.ORG] *On Behalf Of *Tanner, Doug
*Sent:* Monday, October 20, 2014 4:32 PM
*To:* arslist@ARSLIST.ORG
*Subject:* Re: BMC Atrium Core 8.1.02 Installer
**
Yes, Check your SHARE:Application Properties entries and look for a
failed install. Doug
*From:* Action Request System discussion list(ARSList) [mailto:
arslist@ARSLIST.ORG] *On Behalf Of *BradRemedy
*Sent:* Monday, October 20, 2014 6:49 AM
*To:* arslist@ARSLIST.ORG
*Subject:* BMC Atrium Core 8.1.02 Installer
**
Hi
I am trying to upgrade my CMDB 8.1.01 to SP2 and I am getting the
following error:
· [image: Image removed by sender.] You cannot install or
upgrade BMC Atrium CMDB because the installer has detected that a previous
installation failed or is currently running.
Contact BMC Support to resolve the issue.
Do not force the installation or upgrade or delete files manually. Doing
so will make the problem worse.
Has anyone experienced this before and if so any advice?
Thanks in Advance
Brad
_ARSlist: Where the Answers Are and have been for 20 years_
This email is subject to certain disclaimers, which may be reviewed via
the following link. http://compass-usa.com/Pages/Disclaimer.aspx.
_ARSlist: Where the Answers Are and have been for 20 years_
_ARSlist: Where the Answers Are and have been for 20 years_
_ARSlist: Where the Answers Are and have been for 20 years_
_ARSlist: Where the Answers Are and have been for 20 years_
___
UNSUBSCRIBE or access ARSlist Archives at www.arslist.org
Where the Answers Are, and have been for 20 years
ADV: MonitorRemedy from Panacea WAS Open question to BMC on troubleshooting without the WUT
Hi Larry and others,
Troubleshooting without the WUT in a Server Group environment is one of the
use-cases the MonitorRemedy product is designed for. Managing load balanced
server group environments can be a complicated and a time consuming process
Typically, Remedy specialists have to trawl around multiple Mid-Tier and AR
Servers in order to ascertain which server or servers are the problematic ones.
Diagnosing issues involves logging in to each server and looking at resource
usage and multiple logs files.
A unique user search feature of MonitorRemedy allows you to identify which
users are connected to which Mid-tier and AR Servers – allowing specialists to
quickly home in on problematic Mid-tier or AR Servers.
The MonitorRemedy product also provides a specialist graphical view of your
application's health focused on your BMC Remedy AR System and Application
installations.
A graphical warning system lets you quickly identify which Application
function, AR System, Mid-tier or Load Balancer components of your application
are either down or trending towards a potential issue. Drill down features for
example allow you to see:
· If key application components are working (e.g. approvals, email)
· Trends on resource usage , standard query and submit times.
· Log files of all your servers
· Real time ARSystem and Application licence usage.
Setup time for monitoring tools can be time consuming. MonitorRemedy is a
dynamic web application easily deployed to an Apache Tomcat server. The
configuration interface allows you to define the additional applications to be
monitored in minutes, allowing rapid product deployment.
BMC Engage 2014 pricing applies until 17th November 2014 and is £5k GBP per
environment monitored.
You can learn more online at http://www.pws-europe.com/products/MonitorRemedy
Thanks,
Deepak Thohan
cid:image001.gif@01C7FECB.9C4B3860
deepak.tho...@pws-europe.com
_
Telephone: +44 1784 497 047 | Mobile: +44 7875 009519
Fax: +44 1784 497 048 | Skype: deepak.thohan
http://www.pws-europe.com/products/MonitorRemedy
http://www.pws-europe.com/products/ShareRemedy
http://www.pws-europe.com/products/Workflow%20Studio
ShareRemedy | Leverage your investment in Remedy and SharePoint
AnalyseRemedy | The unique Migration, Analysis and Maintenance Tool
http://www.pws-europe.com/products/MobileRemedy MonitorRemedy | BMC Remedy
Environment Monitoring Solution
MobileRemedy| BMC Remedy On Your Mobile
On Wed, Oct 22, 2014 at 5:17 PM, Larry Robinson n...@ncsu.edu wrote:
**
Hi Folks,
John Baker was kind enough to let me know that the script I posted had some
potential security vulnerabilities. He provided the following alternative:
%@ page import=java.net.* %
htmlhead/headbodyp
%
String myhostname= null;
try {
myhostname= InetAddress.getLocalHost().getHostName();
} catch (UnknownHostException e) {
}
if (myhostname!=null) { %
This application server is running on hostname %= myhostname %.
% } %
/p/body/html
Works great. Thanks John!
Larry
On Thu, Oct 16, 2014 at 11:44 PM, Larry Robinson n...@ncsu.edu wrote:
Hi William,
I too receive problem reports from our Mid-tier users and struggled to
determine which server they were connected to. I asked one of my JSP experts to
write us a program that could run from the browser that would emit the name and
IP of the Tomcat server that was serving the users session. We use LVS as our
load balancer, which is IP based so once you are locked onto a server, you stay
there.
I put this JSP program in the Tomcat webapps/ROOT directory and ask the user to
invoke it as:
https://myservicename/server_id.jsp
and have them tell me which server they are locked onto.
Here is the code:
server_id.jsp:
%@ page import=java.util.*,java.io.* %
%
String cmd = hostname;
String outstr = ;
try {
Runtime rt = Runtime.getRuntime();
Process p = rt.exec(cmd);
try {
InputStreamReader ise = new InputStreamReader(p.getErrorStream());
BufferedReader bre = new BufferedReader(ise);
InputStreamReader iso = new InputStreamReader(p.getInputStream());
BufferedReader bro = new BufferedReader(iso);
String line=null;
while ( (line = bre.readLine()) != null ) {
outstr += line;
}
while ( (line = bro.readLine()) != null ) {
outstr += line;
}
}
catch (IOException ioe) {
ioe.printStackTrace();
}
int exitVal = p.waitFor();
}
catch (Throwable t) {
t.printStackTrace();
}
%
br
Hostname: nbsp;nbsp;nbsp; strong%=outstr%/strongbr
Your IP:
nbsp;nbsp;nbsp;nbsp;nbsp;nbsp;nbsp;strong%=request.getRemoteAddr()%/strongbr
Service IP: nbsp;nbsp;nbsp;strong%=request.getLocalAddr()%/strongbr
Hope this is helpful.
Larry
On Thu, Oct 16, 2014 at 1:02 PM, William Rentfrow wrentf...@stratacominc.com
wrote:
**
I sent this to our premiere
Re: Remedy 8.1.2 on VM
Hi Doug - Thanks for the answer - we've been through this with support many times. The question I've never really gotten answered in an authoritative way is this: What exactly does BMC expect us to do with VM to have a local file system when we are using giant enterprise class servers? In the most simple VM setup (not ours) a person might have VM Ware with a file holding the entire VM. This file sits on one (or more depending on RAID, etc) drive of the server. That's not us. We're running VM's off of HP ProLiant BL685c G7's (64 CPU with 524GB Memory). Each VM server is allocated 4 processors and 15GB of RAM (at this point). All of the storage for the applications is ESX storage mounted over NAS. There really isn't any local storage. How do other people deal with this? William Rentfrow wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Wednesday, October 22, 2014 2:31 PM To: arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, It feels like what you are seeing is not a VM vs. physical issue, but an issue about how FTS works and the requirements about configuration to allow scaling and proper behavior. FTS is very FILE SYSTEM intensive. All the indexing of the files and the searching of the indexes is file system dependent. So, what is important for FTS to scale and properly handle large volumes is to be on a LOCAL file system rather than a network file system for the main FTS index directory. This is important whether you are on a physical system or a virtual system. So, there should not be any issue with being virtual, just you need to be virtual tied to a specific physical system that has local disc that has the FTS index directory for full capability. This does constrain the virtual configuration to not just allowing the image to move around to any random machine. This would not apply to all servers, just to the ones that are FTS index servers so that they stay tied to their local file store. Doug Mueller From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Wednesday, October 22, 2014 10:40 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Are you guys who are running VM's using Full Text Search? We've never really gotten that working. We always had the same issue - we'd index stuff, it would work, and the indexes would get corrupted within a few days. BMC wanted us to go to independent physical drives but we have all VM's. We did have separate ESX storage for the FTS directories, shared by all servers. We tried a number of alternative FTS plugin configurations but always ran into some type of issue. It's worth noting this is a high-volume environment. I'd love to know if anyone is using FTS on a VM, and how it is configured (especially with a server group). William Rentfrow wrentf...@stratacominc.commailto:wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Wednesday, October 22, 2014 12:21 PM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Tommy, With all the caveats about making sure you are correctly configuring your VMs and you have properly given them sufficient resources and you have properly done all the right things with how the network and disc and everything is interacting with VMs (all things that should be done regardless of what you are putting in a VM)... There is no issue with running mid-tier and server and any component of the AR System environment or apps on VMs. The DB layer can run in a VM. We have seen evidence of large scale users with lots of data getting better overall throughput on physical machines for the DB, but it will work in both places. BMC itself runs with a physical DB and all AR System servers and mid-tiers on VMs. This is a common configuration in many of our customer environments. I hope this helps, Doug Mueller From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Tommy Morris Sent: Wednesday, October 22, 2014 9:04 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Remedy 8.1.2 on VM ** Just wondering how many are running the current ARS, CMDB, ITSM on VM's. My current hardware is at end of life and I need justification to purchase a new server. Our mid-tiers are running VM with no problem of course but I am concerned about the stability and performance of running ARS and CMDB on a VM instance. Anyone have pros/ cons of going virtual? [cid:image001.png@01CE87C8.A8D7D890] Tommy Morris Sr. Remedy Developer RadioShack Corporation 300 RadioShack Circle Fort Worth, TX 76102-1964 O 817.415.2510
Re: Remedy 8.1.2 on VM
William, The challenge is the performance and reliability of the file system with heavy use. We have found that the heavy interaction with the file system that FTS performs runs into challenges with remote file systems. You can configure servers with local file systems vs. network file systems that are then used to host VMs. You can configure that certain VMs are run on certain machines. So, what customers do is have some machines in their environments with local disc interaction and they configure servers that are FTS indexers to run their VMs on the machines with local file systems. This is still using VMs, but with some specific configurations. For servers not FTS indexers, there is no restriction of local file systems. Not having local file system and using FTS will work - but at a notable performance penalty. So, how do customers deal with it? - don't use FTS or MFS (fewer customers in this camp every day) - accept the performance issue with MFS searches and use FTS sparingly for key large text/attachment fields so even with performance overhead, is faster than DB search - Lock FTS indexer server VMs to specific machines that have local file stores Doug From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Thursday, October 23, 2014 10:11 AM To: arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Hi Doug - Thanks for the answer - we've been through this with support many times. The question I've never really gotten answered in an authoritative way is this: What exactly does BMC expect us to do with VM to have a local file system when we are using giant enterprise class servers? In the most simple VM setup (not ours) a person might have VM Ware with a file holding the entire VM. This file sits on one (or more depending on RAID, etc) drive of the server. That's not us. We're running VM's off of HP ProLiant BL685c G7's (64 CPU with 524GB Memory). Each VM server is allocated 4 processors and 15GB of RAM (at this point). All of the storage for the applications is ESX storage mounted over NAS. There really isn't any local storage. How do other people deal with this? William Rentfrow wrentf...@stratacominc.commailto:wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Wednesday, October 22, 2014 2:31 PM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, It feels like what you are seeing is not a VM vs. physical issue, but an issue about how FTS works and the requirements about configuration to allow scaling and proper behavior. FTS is very FILE SYSTEM intensive. All the indexing of the files and the searching of the indexes is file system dependent. So, what is important for FTS to scale and properly handle large volumes is to be on a LOCAL file system rather than a network file system for the main FTS index directory. This is important whether you are on a physical system or a virtual system. So, there should not be any issue with being virtual, just you need to be virtual tied to a specific physical system that has local disc that has the FTS index directory for full capability. This does constrain the virtual configuration to not just allowing the image to move around to any random machine. This would not apply to all servers, just to the ones that are FTS index servers so that they stay tied to their local file store. Doug Mueller From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Wednesday, October 22, 2014 10:40 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Are you guys who are running VM's using Full Text Search? We've never really gotten that working. We always had the same issue - we'd index stuff, it would work, and the indexes would get corrupted within a few days. BMC wanted us to go to independent physical drives but we have all VM's. We did have separate ESX storage for the FTS directories, shared by all servers. We tried a number of alternative FTS plugin configurations but always ran into some type of issue. It's worth noting this is a high-volume environment. I'd love to know if anyone is using FTS on a VM, and how it is configured (especially with a server group). William Rentfrow wrentf...@stratacominc.commailto:wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Wednesday, October 22, 2014 12:21 PM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Tommy, With all the caveats about making sure you are correctly configuring your VMs and
AREA LDAP and SSL3.0/POODLE
Apologies if this has been answered and/or brought up before. Does ARS 8.1 AREA LDAP use SSL3.0 when making calls to Active Directory? I ask because the infrastructure guys are rolling out a series of POODLE fixes and I need to know if this will break anything. Thanks, Keith Sinclair Remedy Development ShopperTrak Chicago USA O: 312.676.8289 | M: 630.946.4744 ksincl...@shoppertrak.commailto:ksincl...@shoppertrak.com | @shoppertrak www.shoppertrak.comhttp://www.shoppertrak.com/ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Where the Answers Are, and have been for 20 years
CMDB Table
Hi All, Does anyone know what reason there would be to store printer CIs in the BMC.CORE:BMC_ComputerSystem table - with the Class Id as BMC_PRINTER? To my knowledge, printers are stored in the BMC.CORE:BMC_Printers table to inherit the properties of printers. ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Where the Answers Are, and have been for 20 years
Re: CMDB Table
You could have lazy staff like ours. It is too much bother for them to go to two different forms to get information. So just about everything we use is in the BMC.CORE:BMC_ComputerSystem form. Hardware / End User Device /Printer Scanner Monitor Desktop Notebook Etc... Crazy! I know. At least they are updating the assets! From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Kathy Morris Sent: Thursday, October 23, 2014 1:03 PM To: arslist@ARSLIST.ORG Subject: CMDB Table ** Hi All, Does anyone know what reason there would be to store printer CIs in the BMC.CORE:BMC_ComputerSystem table - with the Class Id as BMC_PRINTER? To my knowledge, printers are stored in the BMC.CORE:BMC_Printers table to inherit the properties of printers. _ARSlist: Where the Answers Are and have been for 20 years_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Where the Answers Are, and have been for 20 years
Re: Remedy 8.1.2 on VM
William, Down in this email chain you said : ... We always had the same issue - we'd index stuff, it would work, and the indexes would get corrupted within a few days. ... We did have separate ESX storage for the FTS directories, shared by all servers. What do you mean by shared by all servers? Does each server in the server group have an FTS plugin pointing to a shared collection directory? What version of AR are you running? Keith From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Thursday, October 23, 2014 10:42 AM To: arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, The challenge is the performance and reliability of the file system with heavy use. We have found that the heavy interaction with the file system that FTS performs runs into challenges with remote file systems. You can configure servers with local file systems vs. network file systems that are then used to host VMs. You can configure that certain VMs are run on certain machines. So, what customers do is have some machines in their environments with local disc interaction and they configure servers that are FTS indexers to run their VMs on the machines with local file systems. This is still using VMs, but with some specific configurations. For servers not FTS indexers, there is no restriction of local file systems. Not having local file system and using FTS will work - but at a notable performance penalty. So, how do customers deal with it? -don't use FTS or MFS (fewer customers in this camp every day) -accept the performance issue with MFS searches and use FTS sparingly for key large text/attachment fields so even with performance overhead, is faster than DB search -Lock FTS indexer server VMs to specific machines that have local file stores Doug From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Thursday, October 23, 2014 10:11 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Hi Doug - Thanks for the answer - we've been through this with support many times. The question I've never really gotten answered in an authoritative way is this: What exactly does BMC expect us to do with VM to have a local file system when we are using giant enterprise class servers? In the most simple VM setup (not ours) a person might have VM Ware with a file holding the entire VM. This file sits on one (or more depending on RAID, etc) drive of the server. That's not us. We're running VM's off of HP ProLiant BL685c G7's (64 CPU with 524GB Memory). Each VM server is allocated 4 processors and 15GB of RAM (at this point). All of the storage for the applications is ESX storage mounted over NAS. There really isn't any local storage. How do other people deal with this? William Rentfrow wrentf...@stratacominc.commailto:wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Wednesday, October 22, 2014 2:31 PM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, It feels like what you are seeing is not a VM vs. physical issue, but an issue about how FTS works and the requirements about configuration to allow scaling and proper behavior. FTS is very FILE SYSTEM intensive. All the indexing of the files and the searching of the indexes is file system dependent. So, what is important for FTS to scale and properly handle large volumes is to be on a LOCAL file system rather than a network file system for the main FTS index directory. This is important whether you are on a physical system or a virtual system. So, there should not be any issue with being virtual, just you need to be virtual tied to a specific physical system that has local disc that has the FTS index directory for full capability. This does constrain the virtual configuration to not just allowing the image to move around to any random machine. This would not apply to all servers, just to the ones that are FTS index servers so that they stay tied to their local file store. Doug Mueller From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Wednesday, October 22, 2014 10:40 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Are you guys who are running VM's using Full Text Search? We've never really gotten that working. We always had the same issue - we'd index stuff, it would work, and the indexes would get corrupted within a few days. BMC wanted us to go to independent physical drives but we have all VM's. We did have separate ESX storage for the FTS directories, shared by all servers. We tried a number of alternative FTS plugin
Re: Remedy 8.1.2 on VM
The one we've had the most trouble with is 7.6.04 SP3. And yes - the FTS indexes were created and stored on one mount point which was mounted on all of the AR Servers. The admin server was the only server capable of writing/updating the indexes but all of the servers needed access. We also tried a few different configurations - for example, we tried it with custom FTS plugin configurations so the AR servers were making calls to the admin server instead of reading the files directly - the thought here was that file locking was messing things up, so having only one server touching the files might work. That didn't fix it either. We have never had it be stable for longer than a month - and that was on 7.6.03x. On 7.6.04 we couldn't get it to be stable for more than a few days. We are now working on a couple separate instances of 8.1x and I'm hoping to get it working there. I haven't started yet though. Before I got started I wanted to see how everyone else was doing this in server groups. Also, the first 8.1x server group we are working on has only two servers, so my expectations are that any collision/competition for resources should be much lower. William Rentfrow wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Odom, Keith Sent: Thursday, October 23, 2014 1:58 PM To: arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, Down in this email chain you said : ... We always had the same issue - we'd index stuff, it would work, and the indexes would get corrupted within a few days. ... We did have separate ESX storage for the FTS directories, shared by all servers. What do you mean by shared by all servers? Does each server in the server group have an FTS plugin pointing to a shared collection directory? What version of AR are you running? Keith From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Thursday, October 23, 2014 10:42 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, The challenge is the performance and reliability of the file system with heavy use. We have found that the heavy interaction with the file system that FTS performs runs into challenges with remote file systems. You can configure servers with local file systems vs. network file systems that are then used to host VMs. You can configure that certain VMs are run on certain machines. So, what customers do is have some machines in their environments with local disc interaction and they configure servers that are FTS indexers to run their VMs on the machines with local file systems. This is still using VMs, but with some specific configurations. For servers not FTS indexers, there is no restriction of local file systems. Not having local file system and using FTS will work - but at a notable performance penalty. So, how do customers deal with it? -don't use FTS or MFS (fewer customers in this camp every day) -accept the performance issue with MFS searches and use FTS sparingly for key large text/attachment fields so even with performance overhead, is faster than DB search -Lock FTS indexer server VMs to specific machines that have local file stores Doug From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Thursday, October 23, 2014 10:11 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Hi Doug - Thanks for the answer - we've been through this with support many times. The question I've never really gotten answered in an authoritative way is this: What exactly does BMC expect us to do with VM to have a local file system when we are using giant enterprise class servers? In the most simple VM setup (not ours) a person might have VM Ware with a file holding the entire VM. This file sits on one (or more depending on RAID, etc) drive of the server. That's not us. We're running VM's off of HP ProLiant BL685c G7's (64 CPU with 524GB Memory). Each VM server is allocated 4 processors and 15GB of RAM (at this point). All of the storage for the applications is ESX storage mounted over NAS. There really isn't any local storage. How do other people deal with this? William Rentfrow wrentf...@stratacominc.commailto:wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Wednesday, October 22, 2014 2:31 PM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, It feels like what you are seeing is not a VM vs. physical issue, but an issue about how FTS works and the requirements about configuration to allow scaling and proper
Re: CMDB Table
Kathy, BMC_Printer is a sub-class of BMC_ComputerSystem, which means it (bmc_printer) is a join of BMC_ComputerSystem and BMC_Printer_ (ending with an underscore). Any BMC_Printers will show up in BMC.CORE:BMC_ComputerSystem as well, just with a different ClassId. Hopefully that helps. Here's a link to the common data model for 8.1: https://docs.bmc.com/docs/download/attachments/485372857/403793_CMDB8.1_CDM_Diagram.pdf Thad On Thu, Oct 23, 2014 at 11:18 AM, Sanford, Claire claire.sanf...@memorialhermann.org wrote: ** You could have lazy staff like ours. It is too much bother for them to go to two different forms to get information. So just about everything we use is in the BMC.CORE:BMC_ComputerSystem form. Hardware / End User Device /Printer Scanner Monitor Desktop Notebook Etc… Crazy! I know. At least they are updating the assets! *From:* Action Request System discussion list(ARSList) [mailto: arslist@ARSLIST.ORG] *On Behalf Of *Kathy Morris *Sent:* Thursday, October 23, 2014 1:03 PM *To:* arslist@ARSLIST.ORG *Subject:* CMDB Table ** Hi All, Does anyone know what reason there would be to store printer CIs in the BMC.CORE:BMC_ComputerSystem table – with the Class Id as BMC_PRINTER? To my knowledge, printers are stored in the BMC.CORE:BMC_Printers table to inherit the properties of printers. _ARSlist: Where the Answers Are and have been for 20 years_ _ARSlist: Where the Answers Are and have been for 20 years_ ___ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Where the Answers Are, and have been for 20 years
Re: Remedy 8.1.2 on VM
We no longer recommend the shared mount point configuration. In fact in 7604SP5 and 81SP1, FTS has been reconfigured to have 2 FTS plugins on the indexing server - 1 for the indexing AR Server and 1 for the other servers (readers) in the group. This is the recommended configuration for any version. This does not resolve the slow access time with remote disk systems, but we have not seen index corruption problems since moving to this configuration. As Doug indicated, the simplest and most reliable way to resolve the slow performance of the search engine is to use a local disk. Keith From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Thursday, October 23, 2014 12:29 PM To: arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** The one we've had the most trouble with is 7.6.04 SP3. And yes - the FTS indexes were created and stored on one mount point which was mounted on all of the AR Servers. The admin server was the only server capable of writing/updating the indexes but all of the servers needed access. We also tried a few different configurations - for example, we tried it with custom FTS plugin configurations so the AR servers were making calls to the admin server instead of reading the files directly - the thought here was that file locking was messing things up, so having only one server touching the files might work. That didn't fix it either. We have never had it be stable for longer than a month - and that was on 7.6.03x. On 7.6.04 we couldn't get it to be stable for more than a few days. We are now working on a couple separate instances of 8.1x and I'm hoping to get it working there. I haven't started yet though. Before I got started I wanted to see how everyone else was doing this in server groups. Also, the first 8.1x server group we are working on has only two servers, so my expectations are that any collision/competition for resources should be much lower. William Rentfrow wrentf...@stratacominc.commailto:wrentf...@stratacominc.com Office: 715-204-3061 or 701-232-5697x25 Cell: 715-498-5056 From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Odom, Keith Sent: Thursday, October 23, 2014 1:58 PM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, Down in this email chain you said : ... We always had the same issue - we'd index stuff, it would work, and the indexes would get corrupted within a few days. ... We did have separate ESX storage for the FTS directories, shared by all servers. What do you mean by shared by all servers? Does each server in the server group have an FTS plugin pointing to a shared collection directory? What version of AR are you running? Keith From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of Mueller, Doug Sent: Thursday, October 23, 2014 10:42 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** William, The challenge is the performance and reliability of the file system with heavy use. We have found that the heavy interaction with the file system that FTS performs runs into challenges with remote file systems. You can configure servers with local file systems vs. network file systems that are then used to host VMs. You can configure that certain VMs are run on certain machines. So, what customers do is have some machines in their environments with local disc interaction and they configure servers that are FTS indexers to run their VMs on the machines with local file systems. This is still using VMs, but with some specific configurations. For servers not FTS indexers, there is no restriction of local file systems. Not having local file system and using FTS will work - but at a notable performance penalty. So, how do customers deal with it? -don't use FTS or MFS (fewer customers in this camp every day) -accept the performance issue with MFS searches and use FTS sparingly for key large text/attachment fields so even with performance overhead, is faster than DB search -Lock FTS indexer server VMs to specific machines that have local file stores Doug From: Action Request System discussion list(ARSList) [mailto:arslist@ARSLIST.ORG] On Behalf Of William Rentfrow Sent: Thursday, October 23, 2014 10:11 AM To: arslist@ARSLIST.ORGmailto:arslist@ARSLIST.ORG Subject: Re: Remedy 8.1.2 on VM ** Hi Doug - Thanks for the answer - we've been through this with support many times. The question I've never really gotten answered in an authoritative way is this: What exactly does BMC expect us to do with VM to have a local file system when we are using giant enterprise class servers? In the most simple VM setup (not ours) a person might have VM Ware with a file holding the entire VM. This file sits on one (or more depending on RAID, etc) drive of the server. That's
