我正在尝试让一些多集群事物在我们的两个 GPFS 集群之间工作。一种是存储集群(例如)gpfs01
,另一种是计算集群(例如)gpfs02
。
gpfs02
在我创建mmremotecluster
并正确配置的“ ”集群中mmremotefs
,但是在尝试挂载时,出现以下陈旧文件句柄错误:
[root@gpfs02 ~]# mmmount all
Thu Dec 12 18:57:14 IST 2019: mmmount: Mounting file systems ...
mount: mount gpfs02 on /gpfs/storage failed: Stale file handle.
但尝试mmlsmount
我得到这个:
[root@gpfs02 ~]# mmlsmount all -L
File system gpfs02 (gpfs01:gpfs01) is mounted on 1 nodes:
192.168.1.215 gpfs01 gpfs01
[root@gpfs02 ~]# mmlsmount all_remote
File system gpfs02 (gpfs01:gpfs01) is mounted on 1 nodes.
[root@gpfs02 ~]# tail -f 100 /var/adm/ras/mmfs.log.latest
节目
tail: cannot open '100' for reading: No such file or directory
==> /var/adm/ras/mmfs.log.latest <== 2019-12-12_19:08:47.793+0530: [E] Disk failure. Volume gpfs02. rc = 19. Physical volume nsd3.
2019-12-12_19:08:47.793+0530: [E] Disk failure. Volume gpfs02. rc = 19. Physical volume nsd4.
2019-12-12_19:08:47.794+0530: [X] File System gpfs02 unmounted by the system with return code 19 reason code 0, at line 483 in /project/sprelttn423/build/rttn423s008a/src/avs/fs/mmfs/ts/stripe/stripeopen.C
2019-12-12_19:08:47.794+0530: No such device
2019-12-12_19:08:47.794+0530: Failed to open gpfs02.
2019-12-12_19:08:47.794+0530: No such device
2019-12-12_19:08:47.794+0530: [E] Failed to open gpfs02.
2019-12-12_19:08:47.794+0530: [W] Command: err 666: mount gpfs02
2019-12-12_19:08:47.794+0530: No such device
2019-12-12_19:08:47.855+0530: mmcommon preunmount invoked. File system: gpfs01 Reason: SGPanic
2019-12-12_19:14:21.440+0530: [I] Leaving remote cluster gpfs01
2019-12-12_19:14:21.441+0530: [I] Cluster Manager connection broke. Probing cluster gpfs01
答案1
您必须按照以下步骤创建磁盘:
1) sdb : dd if=/dev/zero of=/sdb count=10240 bs=64k
sdc : dd if=/dev/zero of=/sdc 计数=10240 bs=64k
2)需要创建vim stanza.nsds文件
%nsd: 设备=/sdb
nsd=nsd3
服务器=gpfs01
使用=仅数据
失败组=100
池=数据
%nsd: 设备=/sdc
nsd=nsd4
服务器=gpfs01
使用=仅数据
失败组=100
池=数据
3)之后创建磁盘:
mmcrnsd -F 节.nsds -v 否
3)mmlsnsd
文件系统 磁盘名称 NSD 服务器
(空闲磁盘)nsd3 gpfs01
(空闲磁盘)nsd4 gpfs01
4) 创建文件系统和挂载点。