[cifs-protocol] [EXTERNAL] Re: [MS-DNSP] StartScavenging RPC call - TrackingID#2106010040000444

Jeff McCashland jeffm at microsoft.com
Fri Jun 18 17:05:20 UTC 2021


Hi Douglas,

Just to put a wrap on this issue...

Aging and scavenging applies to individual resource records, and the scavenging process is to delete the record. This triggers a timer on the dnsNode to kick off the tombstoning process. Tombstoned nodes are kept around for a period of time to allow replication the expired status of the node to other servers. 

Please let me know if you have further questions on the StartScavenging RPC call. Otherwise, I'll consider this issue closed. 

Best regards,
Jeff McCashland | Senior Escalation Engineer | Microsoft Protocol Open Specifications Team 
Phone: +1 (425) 703-8300 x38300 | Hours: 9am-5pm | Time zone: (UTC-08:00) Pacific Time (US and Canada)
Local country phone number found here: http://support.microsoft.com/globalenglish | Extension 1138300
We value your feedback.  My manager is Natesha Morrison (namorri), +1 (704) 430-4292

-----Original Message-----
From: Jeff McCashland 
Sent: Wednesday, June 16, 2021 9:26 AM
To: Douglas Bagnall <douglas.bagnall at catalyst.net.nz>; cifs-protocol at lists.samba.org
Cc: Jeff McCashland <jeffm at microsoftsupport.com>
Subject: RE: [EXTERNAL] Re: [MS-DNSP] StartScavenging RPC call - TrackingID#2106010040000444

Resending to get into the list. 

-----Original Message-----
From: Jeff McCashland
Sent: Wednesday, June 9, 2021 11:48 AM
To: Douglas Bagnall <douglas.bagnall at catalyst.net.nz>; cifs-protocol at lists.samba.org
Cc: Jeff McCashland <jeffm at microsoftsupport.com>
Subject: RE: [EXTERNAL] Re: [MS-DNSP] StartScavenging RPC call - TrackingID#2106010040000444

Hi Douglas,

It doesn't appear to me that there is any connection between the Tombstone process and Scavenging/Aging. Surprisingly, these appear to be separate methods of clearing out stale records. In Windows, the processes are handled by separate code that doesn't appear connected. 

Best regards,
Jeff McCashland | Senior Escalation Engineer | Microsoft Protocol Open Specifications Team
Phone: +1 (425) 703-8300 x38300 | Hours: 9am-5pm | Time zone: (UTC-08:00) Pacific Time (US and Canada) Local country phone number found here: http://support.microsoft.com/globalenglish | Extension 1138300 We value your feedback.  My manager is Natesha Morrison (namorri), +1 (704) 430-4292

-----Original Message-----
From: Douglas Bagnall <douglas.bagnall at catalyst.net.nz>
Sent: Tuesday, June 8, 2021 11:40 PM
To: Jeff McCashland <jeffm at microsoft.com>; cifs-protocol at lists.samba.org
Cc: Jeff McCashland <jeffm at microsoftsupport.com>
Subject: [EXTERNAL] Re: [MS-DNSP] StartScavenging RPC call - TrackingID#2106010040000444

thanks Jeff.

I am not quite sure if the second of the two steps you mention is the one described at "DsTombstoneInterval" (under 3.1.1.1.1), or another implementation specific process that precedes that.

If that process is part of "scavenging", the description should reflect that nodes can be tombstoned by RPC calls, so time intervals and aging don't necessarily come into it.


cheers,
Douglas

On 9/06/21 10:05 am, Jeff McCashland wrote:
> Hi Douglas,
> 
> Based on my read of the code, the 'scavenging process' is a combination of a) and b). Initially, resource records that have expired are added to a list of required updates. When a sufficient number of records are expired, then they are removed (deleted) from the database. Also, the delete may be forced by an Admin process or due to Zone state. I don't think we can reliably say how long it will take a record to 'disappear' once it's expired.
> 
> The description of scavenging in the doc is confusing. In the description for 'aging', it mentions "scavenging, a process for cleaning out not-recently used records.", and in:
> 2.2.5.2.4.1 DNS_RPC_ZONE_INFO_W2K
> "Records that have not been refreshed by the expiration of this interval are eligible to be removed during the next scavenging cycle performed by the DNS server."
> 
> Also confusing:
> UpdateScavenge (4 bytes): The number of in-memory nodes modified as a result of a scavenging cycle.
> 
> It seems like we were weakly describing the 2-phase update process without actually describing it. However, I believe this is an implementation detail, and it's up to the server to remove the records immediately or batch the deletes.
> 
> It may be less confusing to say:
> scavenging: A regularly scheduled process in which database records are removed if they have not been updated within a certain time interval, measured by the process that checks whether current time exceeds the record's time stamp value.
> 
> I will file a request to update the document and follow up. Let me know if you have any further questions on scavenging.
> 
> Best regards,
> Jeff McCashland | Senior Escalation Engineer | Microsoft Protocol Open 
> Specifications Team
> Phone: +1 (425) 703-8300 x38300 | Hours: 9am-5pm | Time zone: 
> (UTC-08:00) Pacific Time (US and Canada) Local country phone number 
> found here:
> https://nam06.safelinks.protection.outlook.com/?url=http%3A%2F%2Fsuppo
> rt.microsoft.com%2Fglobalenglish&data=04%7C01%7Cjeffm%40microsoft.
> com%7C87717c3a878d4b524fa308d92b116fba%7C72f988bf86f141af91ab2d7cd011d
> b47%7C1%7C0%7C637588176159674012%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wL
> jAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata
> =OXKH13FBlt9J4MlmvSPTH7FSQca2QwYCJ8dNpJlJT0k%3D&reserved=0 | 
> Extension 1138300 We value your feedback.  My manager is Natesha 
> Morrison (namorri), +1 (704) 430-4292
> 
> -----Original Message-----
> From: Jeff McCashland
> Sent: Tuesday, June 1, 2021 11:07 AM
> To: Douglas Bagnall <douglas.bagnall at catalyst.net.nz>;
> cifs-protocol at lists.samba.org
> Cc: jeffm at microsoftsupport.com
> Subject: [MS-DNSP] StartScavenging RPC call -
> TrackingID#2106010040000444
> 
> [Sree to BCC, support alias on CC, SR ID on Subject]
> 
> Hi Douglas,
> 
> I will research your questions about StartScavenging, and let you know what I find.
> 
> Best regards,
> Jeff McCashland | Senior Escalation Engineer | Microsoft Protocol Open 
> Specifications Team
> Phone: +1 (425) 703-8300 x38300 | Hours: 9am-5pm | Time zone: 
> (UTC-08:00) Pacific Time (US and Canada) Local country phone number 
> found here:
> https://nam06.safelinks.protection.outlook.com/?url=http%3A%2F%2Fsuppo
> rt.microsoft.com%2Fglobalenglish&data=04%7C01%7Cjeffm%40microsoft.
> com%7C87717c3a878d4b524fa308d92b116fba%7C72f988bf86f141af91ab2d7cd011d
> b47%7C1%7C0%7C637588176159674012%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wL
> jAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata
> =OXKH13FBlt9J4MlmvSPTH7FSQca2QwYCJ8dNpJlJT0k%3D&reserved=0 | 
> Extension 1138300 We value your feedback.  My manager is Natesha 
> Morrison (namorri), +1 (704) 430-4292
> 
> -----Original Message-----
> From: Sreekanth Nadendla <srenaden at microsoft.com>
> Sent: Monday, May 31, 2021 8:53 PM
> To: Douglas Bagnall <douglas.bagnall at catalyst.net.nz>;
> cifs-protocol at lists.samba.org
> Subject: 2106010040000444 [MS-DNSP] StartScavenging RPC call
> 
> Dochelp in Bcc
> 
> Hello Douglas, thank you for your question regarding the MS-DNSP specification. We have created incident 2106010040000444  to track the investigation. One of the open specifications team member will contact you shortly.
> 
> Regards,
> Sreekanth Nadendla
> Microsoft Windows Open Specifications
> 
> -----Original Message-----
> From: Douglas Bagnall <douglas.bagnall at catalyst.net.nz>
> Sent: Monday, May 31, 2021 10:45 PM
> To: Interoperability Documentation Help <dochelp at microsoft.com>; 
> cifs-protocol at lists.samba.org
> Subject: [EXTERNAL] [MS-DNSP] StartScavenging RPC call
> 
> hi Dochelp,
> 
> I would like some clarification of the StartScavenging Operation as mentioned in 3.1.4.1 R_DnssrvOperation (and implied in R_DnssrvOperation2, etc).
> 
> The fullest description of the call is:
> 
>> If pszOperation is StartScavenging, the server MUST initiate a resource record scavenging cycle on the DNS server, and return success.
> 
> While in another place (3.1.1.1.1, discussing ScavengingInterval), it 
> says
> 
>> If the value is non-zero and a StartScavenging operation is initiated, the value is ignored and scavenging begins immediately.
> 
> The terms I would like clarification on are "resource record scavenging"
> and "begins immediately".
> 
> Does "resource record scavenging" refer to:
> 
> a) turning stale dynamic records into DNS tombstones when aging is enabled for the zone and server.
> 
> b) removing tombstoned nodes and records after the appropriate time.
> 
> c) both a) and b).
> 
> d) something else?
> 
> 
> Does "begins immediately" imply a scavengable DNS resource should disappear within a few seconds on an unloaded server with a very small zone? Or does the "resource record scavenging cycle" have aspects that mean this is not expected?
> 
> cheers,
> Douglas
> 




More information about the cifs-protocol mailing list