Details
-
Bug
-
Resolution: Incomplete
-
Critical
-
None
-
None
-
lustre 2.10.8 and 2.12.5
mixed OFED, MOFED, Omnipath, and tcp
-
3
-
9223372036854775807
Description
We have two router clusters, which we call RELICs, which connect the infiniband SAN in one building with the infiniband SAN in another building, with ethernet between the routers. All the servers and clients in both buildings, and the router nodes within the clusters which connect to the SAN, are already at lustre 2.12.5. The routers in the RELIC clusters are at lustre 2.10.8. In this configuration, the system is stable.
When we power cycle the RELIC routers and boot them from an image with lustre 2.12.5, the router nodes themselves think everything is fine. There are no errors or warnings on the console, nor indications of failure in the debug log with +net. However, we begin to see symptoms on server nodes which seem to indicate corrupt, dropped, or delayed messages:
LNetError: PPPP:0:(o2iblnd_cb.c:3351:kiblnd_check_txs_locked()) Timed out tx: active_txs, X seconds LNetError: PPPP:0:(o2iblnd_cb.c:3351:kiblnd_check_txs_locked()) Timed out tx: tx_queue, X seconds LNetError: PPPP:0:(o2iblnd_cb.c:3426:kiblnd_check_conns()) Timed out RDMA with ZZZ@o2ib600 (0): c: X, oc: Y, rc: Z LustreError: PPPP:0:(events.c:450:server_bulk_callback()) event type 5, status -103, desc ffffXXXX LustreError: PPPP:0:(ldlm_lib.c:3279:target_bulk_io()) @@@ Reconnect on bulk READ... LustreError: PPPP:0:(ldlm_lib.c:3285:target_bulk_io()) @@@ network error on bulk READ LustreError: PPPP:0:(ldlm_lib.c:3294:target_bulk_io()) @@@ truncated bulk READ 0(1048576) XXX Lustre: PPPP:0:(client.c:2133:ptlrpc_expire_one_request()) @@@ Request sent has failed due to network error: ... Lustre: ls1-OST000e: Bulk IO read error with XXX (at ZZZ@o2ib36), client will retry: rc -110
along with side-affects such as reconnect attempts.
Attachments
Issue Links
- is related to
-
LU-15453 MDT shutdown hangs on mutex_lock, possibly cld_lock
-
- Open
-
Hi Serguei,
> Did you get the chance to experiment with removing related routes prior to upgrading
> the router and adding them back after the upgrade?
This turned out to be harder than I anticipated. We have enough systems and routers that someone there is often one or more routers whose routes have altered to work around some problem being fixed or troubleshot, leading to some difficult corner cases.
At this point we've switched entirely off of Lustre 2.10, and now exclusively use Lustre 2.12 and Lustre 2.15 on the relics. We've seen some of the same symptoms, intermittently, which Gian is investigating. Given that the versions involved are different, and the underlying cause(s) may well be different, I'll close this ticket, and we'll open a new one when there are details.
thanks