About msaravan

msaravan · ‎2011-04-28

Are you using Windows ?? What version of OpsManager are you using ?? Also, can you look at the option reportsArchiveDir and ensure that the path that is mentioned there is exist. "dfm options list reportsArchiveDir" Regards, Saravanan

msaravan · ‎2011-04-28

All the interfaces that are available on the storage system would be discovered when you add storage systems to Operations Manager. Ex: During discovery, for every interface, one of below events would get generated: interface-status:down Error if.adminStatus interface-status:testing Normal if.adminStatus interface-status:unknown Normal if.adminStatus interface-status:up Normal if.adminStatus After that, the interfaces would be monitored continuously. The counter events would get generated only when there is a change. Say, for an example you have an interface e0a on filerA. Case A: If this interface is down, during discovery you will get interface-status:down event generated. In future, if this interface gets up, the counter event interface-status:up would be generated. Case B: If this interface is up, you will get interface-status:up event generated during discovery. In future, if this interface gets down, the counter event interface-status:down would get generated. Thanks, Saravanan

msaravan · ‎2010-06-01

Hi Igor, License should not matter. As long as you see quotas discovered in Operations Manager, that is more than enough for the alerts (e-mails) to be triggered. Regarding your issue, I am having a discussion with my peers. Hope, I'll have some data for you tomorrow. Regards, Saravanan

msaravan · ‎2010-05-28

Hi Igor, I just tested this in my lab setup and I saw that it is working and I got e-mails too. My setup has both DFM Server and Storage system part of same NIS and DNS domains. But, I dont know whether that really matters. This is one of the legacy features of DFM Server and it should work. Since this doesnt work for you, can you file a burt on DFM server product. We need to understand your environment better and some one from Development team can look into your issue. Regards, Saravanan

msaravan · ‎2010-05-27

No, DFM is not using SMTP for storage system communication. DFM uses SNMP and HTTP (NetApp proprietary version) protocols for 95% of its opeartions. For UserQuotas, DFM is using HTTP (NetApp proprietary version) protocol. Regards, Saravanan

msaravan · ‎2010-05-27

On your other question : >> Is there a way to have DFM keep up with the changes in real-time, automatically? Please file a RFE for this. We can make this possible in DFM using SNMP Traps feature. The flow should be : 1. Pre-Requisite : On your storage system, you need to configure DFM server as snmp traphosts (snmp traphost add <DFM_Server_IP>) 2. When DFM server receives traps (quotaExceeded, quotaNormal) from these storage systems, we can have DFM to trigger userquota monitor which will fetch you the latest data. DFM would send mails to the users when their quotas are exceeded. Regards, Saravanan

msaravan · ‎2010-05-27

Hi Igor, Filerview is something like you are on the box directly. So, whatever changes you do on CLI (or via some operations), that will be reflected immediately in Filerview. But, DFM is a Storage Management product. It is designed in a way that the userquotas are monitored only once in 24 hours (1 day). i.e 24 hrs is a default timer set for userquota mon. Again, this option is configurable using "dfm options set". If you reduce this value, it means that DFM will do more polling which in turn will consume few more CPU cycles of your storage system. [root@10 log]# dfm options list | grep -i userQuotaMonInterval userQuotaMonInterval 1 day [root@10 log]# Regards, Saravanan

msaravan · ‎2010-05-26

Hi Igor, I have given step-by-step procedure below which you can use to configure quota alerts for individual users. 1=> Set your email server @ Setup -> Options -> Events and Alerts 2=> Look at the default quota thresholds @ Setup -> Options -> Default Thresholds 3=> Set your e-mail domain-name @ Setup -> Options -> Users. You can leave this field empty if you specify complete e-mail address at User settings page @ step7. 4=> Check that you set the credentials for the storage systems (which have user quotas) being monitored in DFM using dfm host set <storage system> hostlogin=xxxx hostpassword=xxxx. This is needed for the discovery of User quotas. Once the credentials are set, re-discover the storage system using 'dfm host discover <storage_system>'. This is required to keep DFM to have up-to-date information with respect to storage system. 5=> Wait for few mins (till userquotas mon completes) and launch "User Quotas All" Report (Go to Control Center -> Home -> Group Status -> File SRM/Quotas -> Report -> User Quotas, All) 6=> "User Quotal, All" page will show you all the usernames for whom quotas are configured. 7=> Click on one of the Usernames which will launch "User Details" page. Click "Edit Settings" link available in the bottom panel "QuotaUser Tools". It will launch "Edit User Settings" page where you can specify the e-mail address for the user. 8=> You need to repeat the above step for all the users. If you are not comfortable with UI, you can do this using CLI. dfm quota user list dfm quota user set <username> userEmail=abc@xyz.com A simple script might help you here. Hope these steps help you to fix your problem. Regards, Saravanan

msaravan · ‎2010-05-25

Hi lgor, I am not sure about scripts or filer. But we have a product called Operations Manager (part of DFM suite) that can do more than what you expect. The below are the events Operations Manager can generate right today after monitoring your storage system's quota setup: user-disk-space-quota-almost-full user-disk-space-quota-full user-disk-space-quota-ok user-disk-space-soft-limit-exceeded user-disk-space-soft-limit-not-exceeded user-files-quota-almost-full user-files-quota-full user-files-quota-utilization-normal user-files-soft-limit-exceeded user-files-soft-limit-not-exceeded volume-quota-almost-overcommitted volume-quota-not-overcommitted volume-quota-overcommitted Regards, Saravanan

msaravan · ‎2010-05-17

Hi Sridhar Vaddadi, The report "volumes-dedupe-and-lun-file-clones-space-savings" is not available in DFM 3.7.1. Deduplication reports and graphs are enhancement requests addressed only in DFM 3.8. These are made as canned reports and readily available in releases >= 3.8. This report details the space savings of the volumes on which Dedupe is enabled. AFAIK, you cannot customize this report on 3.7.1 since DFM doesnt know the Dedupe status of volumes. Regards, Saravanan

msaravan · ‎2010-05-13

Hi Dolly, As a workaround, if you are looking for filename information, you can create a custom report using SRMFile catalog. This catalog has one of the sub fields called "Name". [root@lnx~]# dfm report catalog list SRMFile SRMFile Catalog Default Display Tab: FileSRM Fields: Field Default Name Default Format ----------------------------------- --------------------------------------------- -------------- All All fields of this catalog Name SRM File Name SRMPath SRM File Path (A) Owner SRM File Owner FileSize SRM File Size B CreateTime SRM File Creation Time DD MMM YYYY 24H AccessTime SRM File Access Time DD MMM YYYY 24H ModificationTime SRM File Modification Time DD MMM YYYY 24H [root@lnx~]# Thanks, Saravanan

msaravan · ‎2010-05-13

As of today, Database views are not available for FSRM. You can try creating equivalent FSRM custom reports. Regards, Saravanan

msaravan · ‎2010-05-10

Hi Kutner, UNC authentication can be set individually for each Host Agents. If you want to run FSRM Walk on UNC path, you have to set 2 things : 1. Login/Password of Hostagent 2. Host Agent CIFS Account / Host Agent CIFS Password The above options can be set both at global level as well as individual Host agent level. To set @ Host Agent level : Go to Home -> Group Status -> File SRM -> Click your Host Agent. It will lauch 'Host Agent Details' page. Click the 'Edit Settings' link available on the bottom left hand side of the page ("Storage Controller Tools"). It'll launch "Edit Host Agent Settings" page. Here, you can set Login/Password, CIFS Account/Password details. Regards, Saravanan

msaravan · ‎2010-04-30

Hi Korns, You can achive this using one of the below ways : 1. Create your SRM path via CLI. In CLI, you can skip the schedule option and it is not mandatory. ex: dfm srm path add <hostname>:<path> [root@lnx ~]# dfm srm path add help NAME add -- list managed SRM paths SYNOPSIS dfm srm path add [ -s <schedule> ] [ -h <agent> ] { all | <object> ... } [root@lnx ~]# 2. In UI also, there is a way you can select "None" as schedule. Goto "Add or Edit SRM Paths" page -> Click "Add A Schedule" -> It will launch "Edit SRM Path Walk Schedules" page. There you can create a new schedule with "Schedule Template" as "None". You can associate this new schedule to the SRM paths which you create. I agree, the UI is little bit tricky but that is how it is as of today. Regards, Saravanan

msaravan · ‎2010-04-28

Hi Terrence, Yes, It is possible. DFM supports this. Refer the FAQ 7.4 & 7.5 at http://now.netapp.com/NOW/knowledge/docs/DFM_win/rel40/html/faq/index.shtml Regards, Saravanan

msaravan · ‎2010-04-19

[Editing the content to remove the sensitive data] I was talking about the bug description there. AFAIK, the issue you reported in this forum is as same as bug 382019. Regards, Saravanan

msaravan · ‎2010-04-19

[Editing the content to remove the sensitive data] After looking at your smmon.log error message, I can say that you hit the issue 382019 undoubtedly. There are 2 cases (case 2 and case 3) talking about this issue. Burt 367643 is a proposal proposed for performance improvement and scale-out solution. Regards, Saravanan

msaravan · ‎2010-03-16

AFAIK, there is no plan to support Host Agent on AIX platforms. If you are looking for this support deliberately, please get in touch with product management team or raise RFEs. Regards, Saravanan

msaravan · ‎2010-02-12

Hi , From your SNMP Walk output, I can understand that node na3c2 has taken over node na3c1. The value of cfSettings.0 at node na3c1 is thisNodeDead(5). This setting indicates that this node has been taken over by its partner who is nothing but na3c2. The one problem I am seeing in your SNMP Walk is that, the cfInterconnectStatus.0 is down @ na3c2. I guess, the interconnect status should be up. So that, the giveback operation which will be executed in future will be succesful. Note : I am not a CF expert. I need to explore more to confirm on this. SNMPv2-SMI::enterprises.789.1.2.3.1.0 = INTEGER: 4 SNMPv2-SMI::enterprises.789.1.2.3.2.0 = INTEGER: 4 SNMPv2-SMI::enterprises.789.1.2.3.3.0 = INTEGER: 1 SNMPv2-SMI::enterprises.789.1.2.3.4.0 = INTEGER: 3 SNMPv2-SMI::enterprises.789.1.2.3.5.0 = Timeticks: (1265719859) 146 days, 11:53:18.59 SNMPv2-SMI::enterprises.789.1.2.3.6.0 = STRING: "na3c1" SNMPv2-SMI::enterprises.789.1.2.3.7.0 = INTEGER: 151704952 SNMPv2-SMI::enterprises.789.1.2.3.8.0 = INTEGER: 2 The interpretation of the above output with their resolvable names is given below : cfSettings.0 = takeoverByPartnerDisabled(4) cfState.0 = takeover(4) cfCannotTakeoverCause.0 = ok(1) cfPartnerStatus.0 = dead(3) cfPartnerLastStatusUpdate.0 = Timeticks: (1265719859) 146 days, 11:53:18.59 cfPartnerName.0 = STRING: "na3c1" cfPartnerSysid.0 = INTEGER: 151704952 cfInterconnectStatus.0 = down(2) Regards, Saravanan

msaravan · ‎2010-02-09

Hi, I guess, you are talking about "cf takeover". If that is the case, then you should look at cf table objects instead of takeoverinfo objects. The objects (takeoverAddrs, takeoverMode, takeoverStatus) you used in your SNMP query belong to NetCache Appliance. If you are looking at the cf settings of your storage system, i.e equivalent to "cf status" command, then you need to query the below objects. 1: cfSettings (1.3.6.1.4.1.789.1.2.3.1) 2: cfState (1.3.6.1.4.1.789.1.2.3.2) 3: cfCannotTakeoverCause (1.3.6.1.4.1.789.1.2.3.3) 4: cfPartnerStatus (1.3.6.1.4.1.789.1.2.3.4) 5: cfPartnerLastStatusUpdate (1.3.6.1.4.1.789.1.2.3.5) 6: cfPartnerName (1.3.6.1.4.1.789.1.2.3.6) 7: cfPartnerSysid (1.3.6.1.4.1.789.1.2.3.7) 8: cfInterconnectStatus (1.3.6.1.4.1.789.1.2.3.8) I pasted the sample walk output of one of my storage systems on cf table (1.3.6.1.4.1.789.1.2.3) below. ***** SNMP QUERY STARTED ***** 1: cfSettings.0 (INTEGER) disabled(3) 2: cfState.0 (INTEGER) cannotTakeover(3) 3: cfCannotTakeoverCause.0 (INTEGER) disabledByOperator(3) 4: cfPartnerStatus.0 (INTEGER) maybeDown(1) 5: cfPartnerLastStatusUpdate.0 (TimeTicks) 146 days 07h:37m:17s.99th (1264183799) 6: cfPartnerName.0 (DisplayString) fas270-184-200 7: cfPartnerSysid.0 (INTEGER) 84189515 8: cfInterconnectStatus.0 (INTEGER) down(2) ***** SNMP QUERY FINISHED ***** Regards, Saravanan

msaravan · ‎2010-02-08

64 bit windows 2008 is one of the certified platforms for Operations Manager 3.8. Please let us know if you hit any issues there. Regards, Saravanan

msaravan · ‎2010-01-18

Even though you follow the workaround by setting diskmoninterval to 1hour 15mins, there is a possibility that you might see duplicate entries in reports for the minimum of 1 hour window. There were few internal burts filed on "disks-multipathing" in Operations Manager. All these burt fixes were already gone into DFM 3.8.1 & DFM 4.0. These DFM versions are having a complete solution for all your disks multipathing problems. Regards, Saravanan

msaravan · ‎2010-01-08

Bad luck !! The MIB doesnt go that deep. Saravanan

msaravan · ‎2010-01-06

I think it is achievable via SNMPWalk. netapp.mib has lot of SNMP objects defined for CIFS. If you are interested in knowing only on the CIFS sessions count per filer, you can use "cifsNSessions" (1.3.6.1.4.1.789.1.7.2.12) object. cifsNSessions OBJECT-TYPE SYNTAX Counter32 MAX-ACCESS read-only STATUS current DESCRIPTION "The current number of active CIFS sessions on the filer." ::= { cifsInfo 12 } Please let me know if you are looking for any other information. I'll go through the SNMP tables and let you know if there is any right SNMP object available for you. Regards, Saravanan

msaravan · ‎2009-12-22

Hi Ray, There are 2 things involved in this query. 1. Generate a report which should list all the spaces in MB/GB instead of KB - For this you can use the method mentioned there. 2. In alerts/events, show the data in MB/GB instead of KB - I think, this is what you are expecting. Am I correct ? Saravanan

Re: Ops Manager and Emailing reports

Re: event 'trigger' definitions? snmp/dfm/opsmngr

Re: Quota notification email

Re: Quota notification email

Re: Quota notification email

Re: Quota notification email

Re: Quota notification email

Re: Quota notification email

Re: Quota notification email

Re: DFM Report volumes-dedupe-and-lun-file-clones-space-savings

Re: Urgent: file SRM database schema documentation

Re: Urgent: file SRM database schema documentation

Re: FSRM UNC path walk in multi-tenant environment

Re: Possible to create an SRM Path without a schedule, or a null schedule?

Re: Can the DFM alarm or event triggered for another action or script to run?

Re: SnapMirror status not updated in Ops Manager

Re: SnapMirror status not updated in Ops Manager

Re: Ops Man Host Agent for AIX?

Re: snmp takeover snmpwalk not working

Re: snmp takeover snmpwalk not working

Re: Operations Manager on 64 bit windows2008

Re: Incorrect number of disks showing on operations manager

Re: CIFS session monitoring with Operations Manager

Re: CIFS session monitoring with Operations Manager

Re: Help me stop the madness: I would like DFM to report in MB or GB not KB..