| Home | Current Systems | Former STK Products | EOL Systems | Components | General Info | Search | Feedback |
| Sun System Handbook - ISO 3.4 June 2011 Internal/Partner Edition | ||
|
|
||
 |
||||||||||||
Solution Type Technical Instruction Sure Solution 1010657.1 : The Proper Way to De-stage Cache in a Sun StorEdge[TM] 351x/33x0 Array is to Use the "shutdown" Command
PreviouslyPublishedAs 214710 Description The Sun StorEdge[TM] 33x0/351x arrays manage cache contents differently depending upon how one resets and issues shutdowns to the controllers during maintenance. This document explains the differences in cache management policies to avoid loss of cache contents and thus customer data. Steps to Follow In the event of a power failure or simply manually powering off the array, the batteries will provide power to the cache for up to 72 hours. When power is restored, cache is then written to disk. The way cache is managed during array resets or shutdowns is different depending upon the management interface being used. Note: Issuing a "Shutdown controller" from either interface covered below will properly flush cache before actually shutting down the controllers and is therefore always a safe operation in regards to cached data. There are two ways to reset the controllers: 1. Via sccli: Any controller reset issued with sccli will properly flush cache to disk before the actual reset of the controllers and is thus a safe way to reset the arrays. For example: sccli> reset controller WARNING: This is a potentially dangerous operation. The controller will go offline for several minutes. Data loss may occur if the controller is currently in use. Are you sure? y sccli: shutting down controller... sccli: controller is shut down sccli: resetting controller... The "WARNING" above is due to the fact that both controllers are reset and therefore all I/O will be halted during the reset. 2. Via curses interface (firmware menu): Any controller reset issued from the firmware menu interface via serial connection WILL NOT write cache to disk before actually resetting the controller. This could result in data corruption if the cache was dirty at the time of the reset. This is especially problematic for these arrays since the firmware interface often prompts the user to reset the controller for changes to take effect. This is done without any warning to the user about about possible data loss. Therefore, to safely reset the controllers from the firmware interface always select "No" when prompted to reset the controller and simply issue a "Shutdown controller" first as follows: From the "Main Menu"
This information is documented in Sun StorEdge[TM] 3000 Family RAID Firmware User's Guide, 817-3711-xx. This behavior is by design and is considered normal. Product Sun StorageTek 3510 FC Array Sun StorageTek 3310 SCSI Array Sun StorageTek 3320 SCSI Array Sun StorageTek 3511 SATA Array Internal Comments This document contains normalized content and is managed by the the Domain Lead(s) of the respective domains. To notify content owners of a knowledge gap contained in this document, and/or prior to updating this document, please contact the domain engineers that are managing this document via the “Document Feedback” alias(es) listed below: storage-os-disk-low-domain@sun.com The Knowledge Work Queue for this article is KNO-STO-VOLUME_DISK
Note: see CR 5031371 reset, minnow, destage, cache, corruption, shutdown, sccli, 3510FC, 3510, 3310, 3320, 3511, audited Previously Published As 74204 Change History Date: 2009-12-24 User Name: sue.copeland@sun.com Action: Currency & Update Date: 2007-06-29 User Name: 7058 Action: Approved Comment: Moved CR reference to internal only section of this doc since it is now contract. Attachments This solution has no attachment |
||||||||||||
|
||||||||||||