This allows mapping an nvme_request back to the
nvme_bdev_io.
This requires bumping up the max number of arguments per
tracepoint. 5 was previously chosen as max since it
exactly fit in 64 bytes (1 cacheline) when all
arguments were stored as uint64_t, but now that we
support uint32_t arguments we can afford extra
arguments when some of them are uint32_t. I've
bumped it to 8 so we can avoid having to touch
this value multiple times if we find some cases
where we need 7 or 8 args.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie2ef5e59d10549860b47542e68c1c34efa63047f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13995
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jacek Kalwas <jacek.kalwas@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
This will allow us to map spdk_bdev_io events
to nvme_request events coming in a future patch.
Since we pass the nvme_bdev_io to the nvme driver
(not the spdk_bdev_io), we need to add tracepoints
for the nvme_bdev_io so that spdk_trace can
do the spdk_bdev_io->nvme_bdev_io->nvme_request
mapping.
An alternative would have been to pass the spdk_bdev_io
as the cb_arg to the nvme driver, but that change
seemed to invasive, and I think we will find other
uses for the nvme_bdev_io events anyways.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Id7519e689b01875093359f41a1ca2af912061a8b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13994
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Jacek Kalwas <jacek.kalwas@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This is just a simple wrapper for now around the
calls to spdk_bdev_io_complete and its
nvme status variant. Upcoming patch will
add an spdk_trace_record to this function as well.
This avoids having to litter spdk_trace_record calls
in too many places.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Id2fb3aeb8b070ad6e09c1dfb9a30a61666a35688
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13993
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Jacek Kalwas <jacek.kalwas@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Community-CI: Mellanox Build Bot
The public interface of lib/accel is now include/spdk/accel.h
Change-Id: Id94f623a494eb1b524b060f4413f633073ea7466
Signed-off-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13916
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
If a nvme_ctrlr is unregistered while I/O path caches are clearing, the
unregistration would fail. This race was not considered for a case that
a nvme_ctrlr is unregsitered when adminq poller started I/O path cache
clearing.
As a bug fix, control ANA log page updating and I/O path cache clearing
separately by adding a new flag io_path_cache_clearing.
Fixes issue #2617
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Id58684caf9b2a10fa68bdb717288ff2bd799c3f7
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13924
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Dong Yi <dongx.yi@intel.com>
This RPC prints only cntlid, to identify a controller
it is needed to issue bdev_get_bdevs and find more info
using cntlid. Printing the controller's trid helps to
identify the controller faster.
Signed-off-by: Alexey Marchuk <alexeymar@nvidia.com>
Change-Id: I97325b822528ef9e71afbe2ff1c30b3bce2ae203
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13655
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Dong Yi <dongx.yi@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
nvme_ctrlr_init_ana_log_page() had used cdata->mnan as active ns count
to allocate a buffer and read a ANA log page into the buffer.
However, cdata->mnan was larger than the real active ns count and caused
an issue when the corresponding NVMe-oF target set the bit 18 of the SGL
support field in the Identify Controller Data structure.
We still need to use cdata->mnan to allocate the buffer because
number of namespaces may be increased dynamically after initialization.
Hence, rename nvme_ctrlr::ana_log_page_size to
nvme_ctrlr::max_ana_log_page_size and calculate and use the active ns
count to read the ANA log page. Check if the current ana_log_page_size
is not larger than nvme_ctrlr->max_ana_log_page_size.
Fixes issue #2584
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ieb10b9c793c4f48ffd88d517c0e9a55184b7d935
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13653
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Calculation of the ANA log page size should use the
identify ctrl MNAN field
(maximum number of allowed namespaces)
not the NN (maximum valid nsid value).
An ANA-enabled controller must have a non-zero MNAN value,
see NVMe Base Specification, Figure 251,
therefore nvme_ctrlr_init_ana_log_page() may safely use MNAN.
Since NN might be much higher than MNAN,
ANA log size based on NN may results in a very large
log page and cause a failure to get ANA log,
e.g. if it is larger than the controller's MDTS.
Fix: replace cdata->nn with cdata->mnan
in nvme_ctrlr_init_ana_log_page()
Signed-off-by: Anton Eidelman <anton@lightbitslabs.com>
Change-Id: I2a522dca833a27dddad25848d7688efa23d23091
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13039
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Jacek Kalwas <jacek.kalwas@intel.com>
Many open source projects have moved to using SPDX identifiers
to specify license information, reducing the amount of
boilerplate code in every source file. This patch replaces
the bulk of SPDK .c, .cpp and Makefiles with the BSD-3-Clause
identifier.
Almost all of these files share the exact same license text,
and this patch only modifies the files that contain the
most common license text. There can be slight variations
because the third clause contains company names - most say
"Intel Corporation", but there are instances for Nvidia,
Samsung, Eideticom and even "the copyright holder".
Used a bash script to automate replacement of the license text
with SPDX identifier which is checked into scripts/spdx.sh.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Iaa88ab5e92ea471691dc298cfe41ebfb5d169780
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12904
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Dong Yi <dongx.yi@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Paul Luse <paul.e.luse@intel.com>
Reviewed-by: <qun.wan@intel.com>
We cannot try to set multipath policy on a non-nvme bdev.
While here, make the error messages match what we use for
setting preferred path.
Fixes issue #2543.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I823077f92634ee3c16e77e7e0d67eb343ec3584e
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12916
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: <qun.wan@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Maximum Data Transfer Size indicates the maximum data transfer size
for a command that transfers data between memory and the controller.
SPDK NVMe driver will split IO request based on MDTS and stripe
boundary, however, if user submited a big enough IO size, SPDK
NVMe driver will run out of requests data structure and return
-EINVAL as the error code, it's OK for application who call SPDK
NVMe driver directly.
SPDK bdev module also will do similar split based on `max_segment_size`
and `max_num_segments`, but we don't set them for bdev/nvme driver,
once there was a big enough IO request, the NVMe driver will return
-EINVAL to bdev module, here we set `max_segment_size` based on MDTS
and set `max_num_segments` based on number of requests of NVMe
controller.
Fix#2403.
Change-Id: Ic6e14a0b12413783597122285ac648b87e1f1e16
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12186
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
During ctrlr reconnection, spdk_nvme_ctrlr_reconnect_poll_async()
is executed by a non-timed poller.
We should poll adminq more often during ctrlr disconnection too.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ib1f5b41015aed20deda8df6f2c837981ac233c04
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12615
Reviewed-by: Dong Yi <dongx.yi@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Michael Haeuptle <michaelhaeuptle@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
As described in the NVMe specification, a controller level reset
includes the following actions:
- the controller stops processing any outstanding admin or I/O
commands;
- all I/O SQs and CQs are deleted.
In a full controller reset sequence for a PCIe controller, if we do
a controller level reset first, we can abort outstanding commands
after the hardware has actually been stopped.
For NVMe-oF controller, each I/O qpair is an independent network
connection and is disconnected safely. We do not want to change
NVMe-oF controller.
Fixes the issue #2360
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: If05febac74705bfd3df5abd15064c1203126e027
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12447
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Michael Haeuptle <michaelhaeuptle@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
It's now possible to specify a time to wait until a connection to the
discovery controller and the NVM controllers it exposes is made.
Whenever that time is exceeded, a callback is immediately executed.
However, depending on the stage of the discovery process, we might need
to wait a while before actually stopping it (e.g. because a controller
attach is in progress). That means that a discovery service might be
visible for a while after it timed out.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I2d01837b581e0fa24c8e777730d88d990c94b1d8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12684
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Currently this should be a no-op as the log_page is always freed in
discovery_remove_controllers(), but it'll make it easier to handle cases
when we want to stop the discovery service while it's attaching NVM
controllers. We'll be relying on this in the subsequent patch adding
attach timeout to start_discovery().
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Ia03fde92bf5ae5590bca507b7a0f963885d85f4f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12721
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Dong Yi <dongx.yi@intel.com>
By default, failback to the preferred I/O path is done automatically
if it is restored. Some users may want to keep using the backup I/O
path even if the preferred I/O path is restored. In this case,
bdev_nvme_set_preferred_path can be used to do manual failback.
We may be able to clear/fill I/O path cache more strictly but it will
be complicated and have bugs. This patch does the minimal change,
just skips an apparent case.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I78fe5faee6ff04e88ae3d7c6be6da1c20637c912
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12431
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Using stop_discovery() will also free the entry contexts tied to that
discovery service. It's not a big deal to leave them, as module_fini is
usually called on app shutdown, but this gets rid of asan reporting
memory leaks.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I81fc843708694791a5846c24cbde4b32e0fa7287
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12683
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
It was never freed, so we leaked it.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I9e17d043ebd0d8d11bb776cb2676ac7f53f8ee41
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12682
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
For now, it's only called from a single location,
bdev_nvme_stop_discovery, but it'll make it possible to stop the
discovery from other places.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I3ba447f30fd8ed23c392d008b3d03cdad30cdb33
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12681
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
If the trid is not zeroed out, trid.trstring might contain garbage
value, which means that nvme_probe_internal() might not populate it
based on trtype and will use that garbage value to get a transport,
leading to the following failure:
```
nvme.c: 834:nvme_probe_internal: *ERROR*: NVMe trtype 3 () not available
```
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I56bc6502d285ee5ce094184e00d3297f6332e8c0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12680
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Moved assigning ctx->start_cb_fn before it's checked for NULL to set
ctx->wait_for_attach, otherwise it was always false.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I819d9e326dbb36f943279c3714695ae604dd64b2
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12628
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Check that we're not already connected to a discovery service that has
the same address (or has a referall to) as the service we're trying to
start.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I84863fd959f62b30e9a348f69d10c7f1edffda7a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12627
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Otherwise, we'll try to use the same name to create the bdevs.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: I38cfb9073f343a7bf3966754008065c7ab89e609
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12626
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
The following patches will swap the ordering of destroying I/O qpairs
and disconnecting a controller for PCIe transport to fix a github issue.
Setting callback and calling spdk_nvme_ctrlr_disconnect() have been
executed in two cases now.
After the following patches, these will be executed in three cases.
Factoring out these into a helper function will be rewarded.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I8597908d7fd8acc6dc62ec442ba2e8c4c08f11a4
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12562
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Michael Haeuptle <michaelhaeuptle@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
The following patches will swap the ordering of destroying I/O qpairs
and disconnecting a controller for PCIe transport to fix a github issue.
For PCIe transport, destroying I/O qpairs will be as the completion
callback to controller disconnection. To do this easier, factor out
destroying I/O qpairs into a helper function.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I96bb6d0487fc67b71c759ee64200bd7fb71aceb1
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12561
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Michael Haeuptle <michaelhaeuptle@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
The RPC returns a list of active discovery service connections. Each
discovery service is described by a name, its trid, and a list of
discovery service trids it refers to.
Signed-off-by: Konrad Sztyber <konrad.sztyber@intel.com>
Change-Id: Ifa4b9501dd353e7b4948ad830575a6c94dafd86b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12380
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
The NVMe bdev module supported active-passive policy for multipath mode
first. By this patch, the NVMe bdev module supports active-active policy
for multipath node next. Following the Linux kernel native NVMe multipath,
the NVMe bdev module supports round robin algorithm for active-active
policy.
The multipath policy, active-passive or active-active, is managed per
nvme_bdev. The multipath policy is copied to all corresponding
nvme_bdev_channels.
Different from active-passive, active-active caches even non_optimized
path to provide load balance across multiple paths.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ie18b24db60d3da1ce2f83725b6cd3079f628f95b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12001
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
If we specify a preferred path manually for each NVMe bdev, we will
be able to realize a simple static load balancing and make the failover
more controllable in the multipath mode.
The idea is to move I/O path to the NVMe-oF controller to the head of
the list and then clear the I/O path cache for each NVMe bdev channel.
We can set the I/O path to the I/O path cache directly but it must be
conditional and make the code very complex. Hence, let find_io_path() do
that.
However, a NVMe bdev channel may be acquired after setting the preferred
path. To cover such case, sort the nvme_ns list of the NVMe bdev too.
This feature supports only multipath mode. The NVMe bdev module supports
failover mode too. However, to support the latter, the new RPC needs to
have trid as parameters and the code and the usage will be come very
complex. Add a note for such limitation.
To verify one by one exactly, add unit test.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ia51c74f530d6d7dc1f73d5b65f854967363e76b0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12262
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: <tanl12@chinatelecom.cn>
Reviewed-by: GangCao <gang.cao@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
ACWU and NACWU are 0-based values. But spdk_bdev_get_acwu()
specifies the compare-and-write-unit in terms of blocks
(i.e. 1-based). So the bdev/nvme module needs to add 1
to this value before registering the bdev.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I7c19975a2bd8c09bb65374838fe20aad690d1ecf
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12384
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
When sending a fused compare and write command, we pass a callback
bdev_nvme_comparev_and_writev_done that we expect to be called twice
before marking the io as completed. In order to detect if a call to
bdev_nvme_comparev_and_writev_done is the first or the second one, we
currently rely on the opcode in cdw0. However, cdw0 may be set to 0,
especially when aborting the command. This may cause use-after-free
issues and this may call the user callbacks twice instead of once.
Use a bit in the nvme_bdev_io instead to keep track of the number of
calls to bdev_nvme_comparev_and_writev_done.
Signed-off-by: Alex Michon <amichon@kalrayinc.com>
Change-Id: I0474329e87648e44b08998d0552b2a9dd5d34ac2
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12180
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Konrad Sztyber <konrad.sztyber@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Add an new RPC bdev_nvme_get_io_paths to query all active I/O paths.
One io_path belongs to One nvme_bdev_channel.
Each nvme_bdev_channel is associated with one nvme_bdev.
If the RPC bdev_nvme_get_io_paths has a bdev name as a parameter
it can use spdk_for_each_channel() simply for the corresponding
nvme_bdev.
However, users will want to know I/O paths of all nvme_bdevs like
the RPC bdev_get_bdevs.
One io_path has one nvme_qpair. One nvme_qpair belongs to one
nvme_poll_group. By relying on these relationships, the RPC
bdev_nvme_get_io_paths traverses all nvme_poll_groups by using
spdk_for_each_channel() to g_bdev_nvme_ctrlrs.
The RPC bdev_nvme_get_io_paths has two modes, display all or
the specified NVMe bdev's active I/O paths.
The specified bdev name is used just for comparison and empty
array is returned if no matched io_path is found.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I4a0dbf3ef7aaa9a7b7345fc03dc493cc6d37bc99
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12146
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
NVMe bdev name already includes the name of the NVMe bdev controller and
the NSID. CNTLID will be a good ID to identify a namespace from a NVMe
bdev when multipath is configured. However, the query RPCs,
bdev_get_bdevs and bdev_nvme_get_controllers had not returned such
information.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I2f2e355ff13f69ced616be803a3152c838cdc980
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12276
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Previously, if a namespace is in ANA inaccessible state, I/O had been
queued infinitely. Fix this issue according to the NVMe spec.
Add a temporary poller anatt_timer and a flag ana_transition_timedout for
each nvme_ns.
Start anatt_timer if the nvme_ns enters ANA transition. If anatt_timer
is expired, set ana_transition_timedout to true. Cancel anatt_timer or
clear ana_transition_timedout if the nvme_ns exits ANA transition.
nvme_io_path_become_available() returns false if ana_transition_timedout
is true.
Add unit test case to verify these addition.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ic76933242046b3e8e553de88221b943ad097c91c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12194
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Monica Kenguva <monica.kenguva@intel.com>
We may fail creating qpair when adding io_path while creating a bdev_channel
if connection is down. But if we enable I/O error recovery, we can retry
creating qpair later.
So let nvme_qpair_create() succeed if the ctrlr is being reset or
I/O error recovery is enabled even if creating qpair failed.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I7d4ff036187bb79ada258cfc299582b4d287018b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12288
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: <tanl12@chinatelecom.cn>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Previously, only if admin qpair gets error, bdev_nvme_failover() was
called.
However, I/O qpair may get error earlier than admin qpair. In this case,
bdev_nvme_failover() was called but reset was already in progress. So
bdev_nvme_failover() returned without doing anything.
bdev_nvme_reset_complete() executes bdev_nvme_failover() if reset
failed. However the test time of test/nvmf/host/failover.sh was very
short. Timeout came before trying bdev_nvme_failover().
We can replace other bdev_nvme_reset() calls by bdev_nvme_failover()
but this patch focuses on the critical case.
Fixes issue #2128.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I68f54bbf54f92343aa56ae41f2b4cd92421c4bbb
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12295
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Setting this optional parameter to true makes the
RPC completion wait until the attach for all
discovered NVM subsystems have completed.
This is especially useful for fio or bdevperf, to
ensure that all of the namespaces are actually
available before testing.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Icf04a122052f72e263a26b3c7582c81eac32a487
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12044
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
After the change that the NVMe bdev module disconnects qpair asynchronously,
disconnected_qpair_cb() got NOTICELOG always when a qpair was disconnected
and freed. This was very noisy.
We have three cases that disconnected_qpair_cb() is called now, 1) qpair
was destroyed in a full ctrlr reset sequence, 2) the upper layer closed
I/O channel, and 3) qpair detected error, and was disconnected and freed.
Get NOTICELOG for 3) but get DEBUGLOG for 1) and 2) with some rewording.
Additionally, to improve readability, change if-else ordering.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ib63bcfd4b72a82a13d3cda208c71cdb40a42fd6b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/12085
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
We want to be able to save the discovery configuration
in a generated JSON-RPC file.
The obvious change needed here is to add a
bdev_nvme_start_discovery RPC to the config file
for each discovery context.
But we also need to make sure we do not emit
bdev_nvme_attach_controller RPCs for controllers
that were attached via the discovery service. These
controllers will be attached by the discovery service
instead - or maybe not at all if the discovery
log page returns different results.
Do both of these changes here, since they are
somewhat tied to each other.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ic2072150c3efdd0a8d01da09e33a647e4929779b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11818
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This keeps track if an nvme_ctrlr was created
implicitly by the discovery service.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I493b7cacfe563737f45a1fffca98855a1929a751
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11817
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
These parameters will be used for any controller created
by the discovery service.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I221b791f38b9c5797ba084c647a98b82c102a121
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11942
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Attempt to start a connection once per second, but
after a connection is successfully started, change
the timer period to one millisecond instead.
This ensures lower response time to AER events when
the discovery controller is operational, but then
decreasing rate of unsuccessful connect attempts (and
associated log messages) if/when a discovery controller
fails.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie24036303f5b00f4a42b6575656f401ea4d578f2
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11774
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Also cycle through the discovery paths if the initial
connect_async() operation fails.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I50f36949d9bba0e3bff81505712076f1a1a7aad5
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11773
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
If a discovery controller fails at some point, we
will want to detach it. This can happen separately
from detaching the controller because we are stopping
the discovery service. So break out the ctrlr
detach operation into a separate phase of the
discovery_poller.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia601b767d32bda1c8899d3a95029781c0aeee136
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11772
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Bdev modules must not access internal bdev_io
structure, so add a new pointer in a public
section. Pointer in internal section will be
used in next patch
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: Ib631563015b3e5fa9300d22b7ae59d8db43c8275
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10421
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Detach and stop are two different operations. This
->detach field was used to denote when the associated
discovery service should be stopped. So call the field
'stop' instead. That may trigger the currently
attached discovery controller to be detached.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I61c7fc860cd9dbcfab71eedfd223c06c51a41f27
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11771
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
We will now keep a list of the possible paths to
the discovery subsystem. One of them will be the
path we are currently connected to (which at service
start, is the path specified by the user).
Additional entries are added for discovery log page
entries referencing the discovery subsystem.
When the discovery service starts, we just have the
initial entry in the list - the discovery poller
tries to connect to it, and if the connect starts
successfully, removes it from the list and points
ctx->entry_ctx_in_use to it.
This will be useful later when we want to iterate
through the available paths to the discovery
subsystem if the current path fails.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I5b18e0f20c4607e29ac0f12f27ba7eb169d0206d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11770
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
This reduces some code duplication since the same
function will be reused in an upcoming patch.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Id6764171ff93c95de49792a4488f2c205b8eddb6
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11769
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
We used to wait until the discovery service could
connect to the discovery subsystem before calling
the callback function provided by the caller (mainly
the start_discovery RPC).
Moving forward, we will be handling the case where
the discovery subsystem is unavailable temporarily.
For now, let's not fail the bdev_nvme_start_discovery
call if we cannot connect to the discovery subsystem.
This will keep the initial service start path the same
as the path where the discovery subsystem is temporarily
unavailable. In the future, we can consider adding
functionality to the start_discovery RPC that waits
up to X number of seconds to see if we were able to
connect and fail otherwise.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Icb05523b9d59f508bfbc0233595c8bf58c10488f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11768
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
For RDMA transport, adminq will find transport error first because
usually only adminq polls CM events.
Change-Id: I7b22cc8883bf02198f1a90d2654c1de6f2e736e6
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11331
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This is a preparation to the following patches.
Change-Id: I1bb0052c745d4f83ff621e4110907a8ac1f1d597
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11330
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
If qpair is disconnected asynchronously, it takes time from detecting
transport error to actually disconnected. We should avoid using the
path as soon as possible after detecting any transport error.
Poll group clears I/O path cache if it finds transport error and avoid
using the path which had transport error.
These changes will reduce the failover time.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I00580159a84372a115ed5e62a6ce13eed4368999
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11329
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
spdk_nvme_ctrlr_disconnect() will be made asynchronous in the
following patches and so we will need to have some changes.
spdk_nvme_ctrlr_disconnect() disconnects adminq and ctrlr synchronously
now.
If spdk_nvme_ctrlr_disconnect() is made asynchronous,
spdk_nvme_ctrlr_process_admin_completions() will complete to disconnect
adminq and ctrlr, and will return -ENXIO only if adminq is disconnected.
However even now spdk_nvme_ctrlr_process_admin_completions() returns
-ENXIO if adminq is disconnected.
So as a preparation, set a callback before calling spdk_nvme_ctrlr_disconnect()
and call the callback if it is set and spdk_nvme_ctrlr_process_admin_completions()
returns -ENXIO.
Besides, fix the return value of bdev_nvme_poll_adminq() in this patch.
Change-Id: I2559f86bb8cf9a92b5b386ed816c00b08c9832df
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10950
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
As we do when deleting ctrlr_channel, disconnect and then free I/O
qpair in a ctrlr reset sequence.
Deleting ctrlr_channel and resetting ctrlr_channel may cause conflicts.
This patch processes such conflicts correctly.
If destroy_ctrlr_channel_cb() is executed between pending and executing
reset_destroy_qpair(), reset_destroy_qpair() is not executed because
ctrlr_channel is not found. In this case, destroy_qpair_channel()
starts disconnecting qpair and deletes ctrlr_channel. Then
disconnected_qpair_cb() releases a reference to poll group.
If destroy_ctrlr_channel_cb() is excuted between executing reset_destroy_qpair()
and disconnected_qpair_cb(), destroy_ctrlr_channel_cb() skips
ctrlr_channel for a reset sequence.
Change-Id: I1f49f74b94aefbea178680aa53ded3a12876c676
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10766
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
When connection is disconnected, bdev_nvme will call
bdev_nvme_failover, and then reset the controller.
nvme_ctrlr->reset_start_tsc should be updated in function
bdev_nvme_failover, then bdev_nvme_check_xxx_timeout can
work well.
Change-Id: I99b639545e9dd4082cdc14696bb7872cb4917b1d
Signed-off-by: Liu Xiaodong <xiaodong.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11957
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
For RDMA transport, current synchronous qpair disconnect occupied CPU for
a second when qpair disconnect gets timeout.
To remove this limitation, we will do the following:
- make spdk_nvme_ctrlr_disconnect_io_qpair() asynchronous,
- spdk_nvme_qpair_process_completions() returns -ENXIO only if the
qpair is actually disconnected.
Even at this patch, spdk_nvme_poll_group_process_completions() invokes
disconnected_qpair_cb only if a qpair is actually disconnected. This
behavior will be maintained.
To use the upcoming asynchronous qpair disconnect easily, when
deleting a ctrlr_channel, disconnect the qpair, and then free the qpair
and release a reference to the poll group when the qpair is actually
disconnected.
We need to delete a nvme_qpair asynchronously after the corresponding
nvme_ctrlr_channel is deleted and defer the deletion of the corresponding
nvme_ctrlr until the nvme_qpair is deleted. To satisfy this requirement,
utilize the reference count of the nvme_ctrlr.
disconnected_qpair_cb() may call spdk_nvme_ctrlr_free_io_qpair() and
spdk_io_device_unregister() successively. The spdk_io_device_unregister()
will execute spdk_nvme_detach_async() from its callback.
spdk_nvme_ctrlr_free_io_qpair() has to complete earlier than
spdk_nvme_detach_async() starts. spdk_nvme_ctrlr_free_io_qpair() is
executed after unwinding stack. spdk_nvme_detach_async() is executed
after sending a message. Sending message is later than unwinding stack.
Hence the requirement is satisfied naturally.
spdk_io_device_unregister() for the nvme_ctrlr is required to be called
on the nvme_ctrlr->thread. To satisfy this requirement, redirect
nvme_ctrlr_unregister() to the nvme_ctrlr->thread. This change is too
small to stand as an independent patch. So include the change in this
patch.
Change-Id: Id8c01966c40b1dae9c4ef17f1b0b3f60a0bd17d5
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10765
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This is another preparation to disconnect qpair asynchronously.
Add nvme_qpair object and move the qpair and poll_group pointers and
the io_path_list list from nvme_ctrlr_channel to nvme_qpair. nvme_qpair
is allocated dynamically when creating nvme_ctrlr_channel, and
nvme_ctrlr_channel points to nvme_qpair.
We want to keep the times of references at I/O path. Change nvme_io_path
to point nvme_qpair instead of nvme_ctrlr_channel, and add
nvme_ctrlr_channel pointer to nvme_qpair.
nvme_ctrlr_channel may be freed earlier than nvme_qpair. nvme_poll_group
lists nvme_qpair instead of nvme_ctrlr_channel and nvme_qpair has a
pointer to nvme_ctrlr.
By using the nvme_ctrlr pointer of the nvme_qpair, a helper function
nvme_ctrlr_channel_get_ctrlr() is not necessary any more. Remove it.
Change-Id: Ib3f579d3441f31b9db7d3844ec56c49e2bb53a5d
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11832
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
The following patches will have the following changes.
Add nvme_qpair object and move qpair and poll_group pointers and
the io_path_list list from nvme_ctrlr_channel to nvme_qpair. nvme_qpair
is allocated dynamically when creating nvme_ctrlr_channel, and
nvme_ctrlr_channel points to nvme_qpair.
qpair is disconnected asynchronously and nvme_ctrlr_channel is
deleted asynchronously.
To make the following patches simpler, refactor two functions,
bdev_nvme_create_ctrlr_channel_cb() and
bdev_nvme_destroy_ctrlr_channel_cb(). The details are as follows.
Factor out nvme_qpair_create() from bdev_nvme_create_ctrlr_channel_cb()
and factor out nvme_qpair_delete() from
bdev_nvme_destroy_ctrlr_channel_cb(). Then reorder a few operation
in these.
Additionally, reorder a operation in _bdev_nvme_add_io_path().
Change-Id: Idf0328fa77a54f40fe52ca72c3842dde82d55972
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11831
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
In the following patches, spdk_nvme_ctrlr_disconnect_io_qpair() will
be changed to be asynchronous, spdk_nvme_ctrlr_disconnect_io_qpair()
will be called first and then spdk_nvme_ctrlr_free_io_qpair() after
the qpair is actually disconnected.
We will not be able to keep the current bdev_nvme_destroy_qpair()
function.
As a preparation, inline bdev_nvme_destroy_qpair() and remove it.
Additionally, this patch has the following changes.
Previously I/O qpair was freed and then I/O path caches were cleared.
Both are SPDK thread local. So there is no dependency for the ordering
of these two operations. However, it will reduce the size of the
following patches if we clear I/O path caches before freeing I/O qpair
when the qpair is disconnected. Hence we clear I/O path caches and then
free I/O qpair.
Remove DTRACE for bdev_nvme_destroy_qpair() for now.
It will be restored in the following patches.
Furthermore, fix potential NULL pointer acces in
bdev_nvme_create_qpair().
Change-Id: I0ab78ccb0d240e56b95b53179341afcd909a31f6
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10746
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Add three options for I/O error resiliency to spdk_nvme_bdev_opts.
Then the RPC bdev_nvme_set_options can configure these.
These can be overridden if these are given by the RPC bdev_nvme_attach_controller.
Change-Id: If3ee23aeef8b7585fe0fb5ec4695df5866fc1e74
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11830
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This is not in the fast path, so using INFOLOG
instead of DEBUGLOG allows these messages to be
enabled in release builds.
While here, set this flag in the discovery.sh
test script so that we get better information if
there are test failures.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I1c0d087b5c0cb40118691f4a1bc16adc2fdaad9c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11932
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
The following patches will enable us to specify I/O error resiliency
options per nvme_ctrlr as global options. To do it easier, move
per controller options about I/O error resiliency into struct nvme_ctrlr_opts.
prchk_flags is not exactly for resiliency but move it into struct
nvme_ctrlr_opts too.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I85fd1738bb6e293cd804b086ade82274485f213d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11829
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
The following patches will add options per struct nvme_ctrlr in the
NVMe bdev module. bdev_opts will be used for it.
Additionally, fabrics_connect_timeout_us is set directly to
spdk_nvme_ctrlr_opts. So remove it from the RPC request structure.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I981cda5e69375edc43a8581cd3b43497c38a3d56
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11827
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
If 0 - UINT32_MAX or UINT32_MAX - 0 is substituted into a int variable,
we cannot get any expected result.
Fix the bug and add unit test case to verify the fix.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ib045273238753e16755328805b38569909c8b83a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11836
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
This will be helpful in later patches, when we handle
detach not just at discovery service stop, but also
when a discovery controller is disconnected.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ie62d62f73b328c6e058f6480c61fbdf91e854e2a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11767
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
If the path to a subsystem changes from one discovery
log to the next, we should add the new paths first,
and only then remove paths. This ensures we don't
remove the last path to a subsystem, causing associated
bdevs to get unregisterd and reregistered.
This requires adding a new log_page member to
discovery_ctx, since we now need to walk the log
page to find removed paths after all the new paths
are attached.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I99fc2e40e6f7e2e26d558ebe7bc5208fe474c0ea
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11766
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
These macros are used to prefix the following to
any discovery-related DEBUGLOG or ERRLOG:
Discovery[127.0.0.1:8009]
Inside the brackets are the traddr and trsvcid of
the discovery service associated with that message.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ib1991a13f550bb8c9aaf1194a56b218cbd71c96c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11733
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This is useful for adding trid details to discovery
related log messages in a later patch.
Future patches will update this trid if the
current discovery ctrlr fails and we need to fail
over to a different path to the discovery subsystem.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I51712bab2d891ae9c683f8716b4228741f64e7db
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11732
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
For now, just allocate entries and put them on a new TAILQ on
the discovery_ctx. Future patches will use these to try
to reattach to the discovery subsystem if the current discovery
ctrlr fails.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I3cd841df2260bbe8a497bbbf36dea4a1081f25c0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11731
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
It will be referenced in a second location in
an upcoming patch, so move its definition now to
reduce the size of that patch and avoid a forward
declaration.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Iae12cc613190c03f0d48d71475df98384f8e47c7
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11730
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
This name better describes the purpose of this structure.
Currently it is used to represent discovery log page entries
for NVM subsystems found by the discovery service. Upcoming
patches will also use this structure to represent discovery
log page entries for the discovery subsystem.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I84996c9968200c50c32427f0233cb707cdc2d54c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11547
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
For now, if the discovery service finds a discovery subsystem,
don't connect to it. Support for nested discovery controllers
will be coming soon, but for now we need to make sure we don't
try to connect to a discovery subsystem as if it was an NVM
subsystem.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I00234718b0e39eda6e1cb1b1150a4fadcf6d8b11
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11546
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
This is a bug fix. free() was called to the object allocated by
spdk_malloc().
Hence
free(): invalid pointer: 0x00002000146ece00
was printed.
This was found during multipath testing.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Icf6aa6dcdda728fef91b3acad7a1f1ee219c27af
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11710
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
In a test case, test/nvmf/host/failover.sh, we got ANA error even if
the target did not enable ANA reporting.
We marked the corresponding namespace as ANA state updating but we had
no way to clear it.
Check if we can read ANA log page before setting the flag.
If read ANA log page failed, disable ANA feature until the nvme_ctrlr
is created again. In this operation, all ana_state_updating flags are
cleared.
Fixes#2335
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I4e2608a35d9dfa0395ad74fceebae9faf8cd973c
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11399
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
It may take a long time to detect network transport error
when e.g. port is removed on remote target. This timeout
depends on 2 parameters - retry_count and ack_timeout.
bdev_nvme_set_options supports configuration of retry_count
but transport_ack_timeout is missed. Note: this parameter
is used by RDMA transport only.
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: I7c3090dc8e4078f64d444e2392a9e0a6ecdc31c0
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11175
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: <tanl12@chinatelecom.cn>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
This patch aligns namespace comparison with Linux kernel
implementation:
- UUID is optional and may be NULL
- command set (CSI) should be the same
Signed-off-by: Evgeniy Kochetov <evgeniik@nvidia.com>
Change-Id: I8f889989f24cd51b104057217f87eb303b30fa68
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11312
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Add more dtrace probes to help with identifying issues
in production.
Change-Id: I8fb621a15c5e33ae94d75b4fc31135e2635dcfce
Signed-off-by: Krzysztof Karas <krzysztof.karas@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10561
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Now that we have a much more robust retry framework,
set the default bdev_retry_count to 3. Users can
still override this default with the bdev_nvme_set_options
RPC as before. This ensures that by default, we will
retry I/O when possible.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I045bf4969d02be32b951e72a148ce6b6e251dec1
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11107
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Dong Yi <dongx.yi@intel.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
If ctrlr_loss_timeout_sec is set to -1, reconnect is tried repeatedly
indefinitely, and I/Os continue to be queued.
This patch adds another option fast_io_fail_timeout_sec, a flag
fast_io_fail_timedout to nvme_ctrlr.
If the time fast_io_fail_timeout_sec passed after starting reset,
set fast_io_fail_timedout to true not to use the path for I/O submission.
fast_io_fail_timeout_sec is initialized to zero as same as
ctrlr_loss_timeout_sec and reconnect_delay_sec.
The name of the parameter follows the famous DM-multipath, its fast_io_fail_tmo.
Change-Id: Ib870cf8e2fd29300c47f1df69617776f4e67bd8c
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10301
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Previously reconnect retry was not controlled and was repeated indefinitely.
This patch adds two options, ctrlr_loss_timeout_sec and reconnect_delay_sec,
to nvme_ctrlr and add reset_start_tsc, reconnect_is_delayed, and
reconnect_delay_timer to nvme_ctrlr to control reconnect retry.
Both of ctrlr_loss_timeout_sec and reconnect_delay_sec are initialized to
zero. This means reconnect is not throttled as we did before this patch.
A few more changes are added.
Change nvme_io_path_is_failed() to return false if reset is throttled
even if nvme_ctrlr is reseting or is to be reconnected.
spdk_nvme_ctrlr_reconnect_poll_async() may continue returning -EAGAIN
infinitely. To check out such exceptional case, use ctrlr_loss_timeout_sec.
Not only ctrlr reset but also non-multipath ctrlr failover is controlled.
So we need to include path failover into ctrlr reconnect.
When the active path is removed and switched to one of the alternative paths,
if ctrlr reconnect is scheduled, connecting to the alternative path is left
to the scheduled reconnect.
If reset or reconnect ctrlr is failed and the retry is scheduled,
switch the active path to one of alternative paths.
Restore unit test cases removed in the previous patches.
Change-Id: Idec636c4eced39eb47ff4ef6fde72d6fd9fe4f85
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10128
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Monica Kenguva <monica.kenguva@intel.com>
This is a clean up as a preparation to the following patches.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ib8bc90e17f52086d4e887463e04f65273bb1079b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11068
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
We noticed the difference between the SPDK 21.10 and the latest master
in a test.
The simplified scenario is as follows:
1. Start SPDK NVMe-oF target
2. Run bdevperf for the target with -f parameter to suppress exit
on failure.
3. Kill the target after I/O started.
With the SPDK 21.10, bdevperf retries failed I/Os and exits after
the test time is over.
With the latest SPDK master, bdevperf hungs and does not exit even
after the test time is over.
The cause was as follows:
reset ctrlr is repeated very quickly (once per 10ms by default) and hence
I/Os were queued infinitely because nvme_io_path_is_failed() returned
false if nvme_ctrlr is resetting.
We should queue I/O when nvme_ctrlr is resetting only if reset is throttoled
and fail-fast for the repeated failures is supported.
Hence in this patch, fix the degradation and remove the related unit
test cases.
Reported-by: Evgeniy Kochetov <evgeniik@nvidia.com>
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I4047d42dc44488a05264c6a841d101a7c371358b
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11062
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
This RPC will stop the specified discovery service,
including detaching from any controllers that were
attached as part of that discovery service.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I9222876457fc45e1acde680a7bd1925917c22308
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10832
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
If a path whose namespace is optimized is restored, the corresponding
I/O path cache should be cleared and the path should be chosen as the
optimal path.
This bug was found by a system test.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Ibc3983dbff3418adb090a09df32c2a92a8910d05
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/11004
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Monica Kenguva <monica.kenguva@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Rename a few functions for a full ctrlr reset sequence to
clarify what we do and make the following patches easier.
Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: I051e3ab68c3cd77fd6040a2d069d50a700123ae6
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10920
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
We sort secondary trids to avoid using disconnected trids for failover.
However the sort had a bug.
This bug was found by running test/nvmf/host/multipath.sh in a loop.
Verify the fix by adding unit test.
Fixes#2300
Signed-off-by: Shuhei Matsumoto <shuheimatsumoto@gmail.com>
Change-Id: I22b0ede4d2ef98b786c3e0d1f5337a2d568ba56d
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10921
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
This patch adds the framework for a discovery
service in the bdev/nvme module.
Users can specify an IP/port of a discovery service.
The bdev/nvme module will connect to a discovery
controller, get the discovery log page, and then
register for AERs. It will connect to each
subsystem specified in the initial log page.
AER completions will trigger fetching the log
page again, at which point new subsystems will
be connected to, or removed subsystems will be
detached.
This patch does the following:
* Adds the new start_discovery RPC
* Connects to the discovery controller
* Gets the discovery log page
* Registers for AERs
* Detach from discovery controllers at shutdown
Subsequent patches in this series will:
* Connect to subsystems listed in discovery log page
* Detach from subsystems that were listed in earlier
discovery log pages but subsequently removed
* Add a stop_discovery RPC
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: I54bfa896a48c5619676f156b5ea9f2d1f886c72f
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10694
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
If qpair creation failed, ctrlr_ch remains in group->ctrlr_ch_list
but memory for ctrlr_ch is freed. Next attempt to get ctrlr's io
channel will modify data in already freed memory and may corrupt
another allocation.
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: I85002f2e6ac86a0ffda6dabfa57e79b59074fb5a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10840
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Shuhei Matsumoto <shuheimatsumoto@gmail.com>
Reviewed-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
It is possible that the application calls
get_io_channel during nvme controller reset.
In that case IO qpair won't be created and the
application will get a NULL pointer.
It is possible to repeat get_io_channel later but
there is no such indiciation for the application,
so it can't distinguish between a real failure and
"try again" case during controller reset.
This patch ignores IO qpair creation error if
controller is resetting. IO qpair will be created
when reset completes.
Signed-off-by: Alexey Marchuk <alexeymar@mellanox.com>
Change-Id: Id39202f5a6878453ff54e35df91d5dc49a5f046a
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10828
Community-CI: Mellanox Build Bot
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuheimatsumoto@gmail.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
We will want to use bdev_nvme_create() to attach to
controllers identified through discovery. In this case,
we won't be reporting bdev names back to an RPC caller,
so there's no need to allocate an array of names to be
filled out since they won't be used.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ia386d034df2c2d5a60f9aa18338ba415ec03d763
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10689
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
We will need to add another step in the fini path for
stopping discovery pollers, so this patch prepares for
that.
Signed-off-by: Jim Harris <james.r.harris@intel.com>
Change-Id: Ifecbbac60262f3aae7f7a7ced09b7a600df7c2e8
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/10590
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@mellanox.com>
